diff --git "a/checkpoint-1026/trainer_state.json" "b/checkpoint-1026/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1026/trainer_state.json"
@@ -0,0 +1,8241 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 18.660550458715598,
+  "eval_steps": 500,
+  "global_step": 1026,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 0.04378490149974823,
+      "learning_rate": 4.999989423013716e-05,
+      "loss": 0.6713,
+      "num_input_tokens_seen": 44136,
+      "step": 1
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.040646992623806,
+      "learning_rate": 4.999957692144361e-05,
+      "loss": 0.533,
+      "num_input_tokens_seen": 83096,
+      "step": 2
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.04658753052353859,
+      "learning_rate": 4.999904807660428e-05,
+      "loss": 0.6048,
+      "num_input_tokens_seen": 122112,
+      "step": 3
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.04322144016623497,
+      "learning_rate": 4.999830770009406e-05,
+      "loss": 0.4948,
+      "num_input_tokens_seen": 163064,
+      "step": 4
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.06536195427179337,
+      "learning_rate": 4.999735579817769e-05,
+      "loss": 0.6607,
+      "num_input_tokens_seen": 203808,
+      "step": 5
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.059904925525188446,
+      "learning_rate": 4.9996192378909786e-05,
+      "loss": 0.5802,
+      "num_input_tokens_seen": 241824,
+      "step": 6
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.19818365573883057,
+      "learning_rate": 4.999481745213471e-05,
+      "loss": 0.5148,
+      "num_input_tokens_seen": 287608,
+      "step": 7
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.05985472351312637,
+      "learning_rate": 4.9993231029486544e-05,
+      "loss": 0.5714,
+      "num_input_tokens_seen": 325320,
+      "step": 8
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.061375778168439865,
+      "learning_rate": 4.999143312438893e-05,
+      "loss": 0.6812,
+      "num_input_tokens_seen": 369848,
+      "step": 9
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.06196414306759834,
+      "learning_rate": 4.998942375205502e-05,
+      "loss": 0.5358,
+      "num_input_tokens_seen": 415104,
+      "step": 10
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.07861393690109253,
+      "learning_rate": 4.9987202929487275e-05,
+      "loss": 0.6527,
+      "num_input_tokens_seen": 467224,
+      "step": 11
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.05596446990966797,
+      "learning_rate": 4.99847706754774e-05,
+      "loss": 0.5354,
+      "num_input_tokens_seen": 502824,
+      "step": 12
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.05289844051003456,
+      "learning_rate": 4.998212701060612e-05,
+      "loss": 0.5263,
+      "num_input_tokens_seen": 544744,
+      "step": 13
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.04996591433882713,
+      "learning_rate": 4.997927195724303e-05,
+      "loss": 0.5536,
+      "num_input_tokens_seen": 591136,
+      "step": 14
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.05822828412055969,
+      "learning_rate": 4.997620553954645e-05,
+      "loss": 0.6106,
+      "num_input_tokens_seen": 629664,
+      "step": 15
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.06353770196437836,
+      "learning_rate": 4.997292778346312e-05,
+      "loss": 0.5129,
+      "num_input_tokens_seen": 663392,
+      "step": 16
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.07256966829299927,
+      "learning_rate": 4.996943871672807e-05,
+      "loss": 0.6377,
+      "num_input_tokens_seen": 698360,
+      "step": 17
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.055458713322877884,
+      "learning_rate": 4.996573836886435e-05,
+      "loss": 0.4083,
+      "num_input_tokens_seen": 737520,
+      "step": 18
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.07792335003614426,
+      "learning_rate": 4.9961826771182784e-05,
+      "loss": 0.6086,
+      "num_input_tokens_seen": 768056,
+      "step": 19
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.06627275049686432,
+      "learning_rate": 4.995770395678171e-05,
+      "loss": 0.4591,
+      "num_input_tokens_seen": 806256,
+      "step": 20
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.05830290913581848,
+      "learning_rate": 4.9953369960546676e-05,
+      "loss": 0.3731,
+      "num_input_tokens_seen": 842336,
+      "step": 21
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.07277437299489975,
+      "learning_rate": 4.9948824819150185e-05,
+      "loss": 0.6243,
+      "num_input_tokens_seen": 876672,
+      "step": 22
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.07477546483278275,
+      "learning_rate": 4.994406857105136e-05,
+      "loss": 0.5788,
+      "num_input_tokens_seen": 915192,
+      "step": 23
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.06912907212972641,
+      "learning_rate": 4.993910125649561e-05,
+      "loss": 0.4753,
+      "num_input_tokens_seen": 951904,
+      "step": 24
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.0655476376414299,
+      "learning_rate": 4.993392291751431e-05,
+      "loss": 0.4518,
+      "num_input_tokens_seen": 1001816,
+      "step": 25
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.06466512382030487,
+      "learning_rate": 4.992853359792444e-05,
+      "loss": 0.5638,
+      "num_input_tokens_seen": 1053064,
+      "step": 26
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.0645688995718956,
+      "learning_rate": 4.99229333433282e-05,
+      "loss": 0.4644,
+      "num_input_tokens_seen": 1086688,
+      "step": 27
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.07181251049041748,
+      "learning_rate": 4.9917122201112656e-05,
+      "loss": 0.6191,
+      "num_input_tokens_seen": 1134824,
+      "step": 28
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.07322589308023453,
+      "learning_rate": 4.9911100220449293e-05,
+      "loss": 0.6752,
+      "num_input_tokens_seen": 1172072,
+      "step": 29
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.06396070122718811,
+      "learning_rate": 4.990486745229364e-05,
+      "loss": 0.3587,
+      "num_input_tokens_seen": 1211096,
+      "step": 30
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.07803395390510559,
+      "learning_rate": 4.989842394938482e-05,
+      "loss": 0.459,
+      "num_input_tokens_seen": 1259456,
+      "step": 31
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.05974648892879486,
+      "learning_rate": 4.989176976624511e-05,
+      "loss": 0.4148,
+      "num_input_tokens_seen": 1306944,
+      "step": 32
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.09784268587827682,
+      "learning_rate": 4.988490495917947e-05,
+      "loss": 0.539,
+      "num_input_tokens_seen": 1353744,
+      "step": 33
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.09906516224145889,
+      "learning_rate": 4.987782958627508e-05,
+      "loss": 0.5453,
+      "num_input_tokens_seen": 1394736,
+      "step": 34
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.08984062820672989,
+      "learning_rate": 4.987054370740083e-05,
+      "loss": 0.468,
+      "num_input_tokens_seen": 1442048,
+      "step": 35
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.08672655373811722,
+      "learning_rate": 4.9863047384206835e-05,
+      "loss": 0.4078,
+      "num_input_tokens_seen": 1478440,
+      "step": 36
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.1327345073223114,
+      "learning_rate": 4.9855340680123905e-05,
+      "loss": 0.5299,
+      "num_input_tokens_seen": 1525992,
+      "step": 37
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.09178602695465088,
+      "learning_rate": 4.9847423660363e-05,
+      "loss": 0.439,
+      "num_input_tokens_seen": 1555608,
+      "step": 38
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.09418320655822754,
+      "learning_rate": 4.983929639191469e-05,
+      "loss": 0.5337,
+      "num_input_tokens_seen": 1597392,
+      "step": 39
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.08294719457626343,
+      "learning_rate": 4.983095894354858e-05,
+      "loss": 0.4536,
+      "num_input_tokens_seen": 1649656,
+      "step": 40
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.09774205833673477,
+      "learning_rate": 4.982241138581273e-05,
+      "loss": 0.5221,
+      "num_input_tokens_seen": 1695952,
+      "step": 41
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.09319107979536057,
+      "learning_rate": 4.9813653791033057e-05,
+      "loss": 0.4279,
+      "num_input_tokens_seen": 1737224,
+      "step": 42
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.09561405330896378,
+      "learning_rate": 4.980468623331273e-05,
+      "loss": 0.5121,
+      "num_input_tokens_seen": 1772320,
+      "step": 43
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.08274025470018387,
+      "learning_rate": 4.979550878853154e-05,
+      "loss": 0.54,
+      "num_input_tokens_seen": 1823888,
+      "step": 44
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.08728913217782974,
+      "learning_rate": 4.9786121534345265e-05,
+      "loss": 0.4488,
+      "num_input_tokens_seen": 1872488,
+      "step": 45
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.0787016749382019,
+      "learning_rate": 4.9776524550184965e-05,
+      "loss": 0.4353,
+      "num_input_tokens_seen": 1924744,
+      "step": 46
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.10952188074588776,
+      "learning_rate": 4.97667179172564e-05,
+      "loss": 0.4784,
+      "num_input_tokens_seen": 1959936,
+      "step": 47
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.08525826781988144,
+      "learning_rate": 4.975670171853926e-05,
+      "loss": 0.3586,
+      "num_input_tokens_seen": 2003896,
+      "step": 48
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.10409987717866898,
+      "learning_rate": 4.9746476038786496e-05,
+      "loss": 0.4451,
+      "num_input_tokens_seen": 2047632,
+      "step": 49
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.0782993957400322,
+      "learning_rate": 4.973604096452361e-05,
+      "loss": 0.3591,
+      "num_input_tokens_seen": 2096928,
+      "step": 50
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.09829951077699661,
+      "learning_rate": 4.9725396584047925e-05,
+      "loss": 0.3415,
+      "num_input_tokens_seen": 2129536,
+      "step": 51
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.10606162995100021,
+      "learning_rate": 4.971454298742779e-05,
+      "loss": 0.3758,
+      "num_input_tokens_seen": 2169144,
+      "step": 52
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.09280356764793396,
+      "learning_rate": 4.97034802665019e-05,
+      "loss": 0.485,
+      "num_input_tokens_seen": 2207720,
+      "step": 53
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.11888203024864197,
+      "learning_rate": 4.9692208514878444e-05,
+      "loss": 0.3469,
+      "num_input_tokens_seen": 2236392,
+      "step": 54
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.13222463428974152,
+      "learning_rate": 4.9680727827934354e-05,
+      "loss": 0.4284,
+      "num_input_tokens_seen": 2259088,
+      "step": 55
+    },
+    {
+      "epoch": 1.018348623853211,
+      "grad_norm": 0.10572745651006699,
+      "learning_rate": 4.966903830281449e-05,
+      "loss": 0.4186,
+      "num_input_tokens_seen": 2298496,
+      "step": 56
+    },
+    {
+      "epoch": 1.036697247706422,
+      "grad_norm": 0.11462350189685822,
+      "learning_rate": 4.965714003843079e-05,
+      "loss": 0.4696,
+      "num_input_tokens_seen": 2333016,
+      "step": 57
+    },
+    {
+      "epoch": 1.0550458715596331,
+      "grad_norm": 0.11215240508317947,
+      "learning_rate": 4.9645033135461494e-05,
+      "loss": 0.3905,
+      "num_input_tokens_seen": 2367992,
+      "step": 58
+    },
+    {
+      "epoch": 1.073394495412844,
+      "grad_norm": 0.0973561555147171,
+      "learning_rate": 4.963271769635024e-05,
+      "loss": 0.3588,
+      "num_input_tokens_seen": 2415328,
+      "step": 59
+    },
+    {
+      "epoch": 1.091743119266055,
+      "grad_norm": 0.10240709036588669,
+      "learning_rate": 4.962019382530521e-05,
+      "loss": 0.5532,
+      "num_input_tokens_seen": 2454792,
+      "step": 60
+    },
+    {
+      "epoch": 1.110091743119266,
+      "grad_norm": 0.0959337130188942,
+      "learning_rate": 4.9607461628298244e-05,
+      "loss": 0.331,
+      "num_input_tokens_seen": 2503072,
+      "step": 61
+    },
+    {
+      "epoch": 1.1284403669724772,
+      "grad_norm": 0.10228750854730606,
+      "learning_rate": 4.9594521213063974e-05,
+      "loss": 0.3728,
+      "num_input_tokens_seen": 2546960,
+      "step": 62
+    },
+    {
+      "epoch": 1.146788990825688,
+      "grad_norm": 0.09403488785028458,
+      "learning_rate": 4.958137268909887e-05,
+      "loss": 0.4695,
+      "num_input_tokens_seen": 2595432,
+      "step": 63
+    },
+    {
+      "epoch": 1.165137614678899,
+      "grad_norm": 0.11396344751119614,
+      "learning_rate": 4.9568016167660334e-05,
+      "loss": 0.3653,
+      "num_input_tokens_seen": 2633912,
+      "step": 64
+    },
+    {
+      "epoch": 1.18348623853211,
+      "grad_norm": 0.09487481415271759,
+      "learning_rate": 4.9554451761765766e-05,
+      "loss": 0.3498,
+      "num_input_tokens_seen": 2680792,
+      "step": 65
+    },
+    {
+      "epoch": 1.2018348623853212,
+      "grad_norm": 0.1249895691871643,
+      "learning_rate": 4.9540679586191605e-05,
+      "loss": 0.4053,
+      "num_input_tokens_seen": 2716584,
+      "step": 66
+    },
+    {
+      "epoch": 1.2201834862385321,
+      "grad_norm": 0.12268221378326416,
+      "learning_rate": 4.952669975747232e-05,
+      "loss": 0.4189,
+      "num_input_tokens_seen": 2757088,
+      "step": 67
+    },
+    {
+      "epoch": 1.238532110091743,
+      "grad_norm": 0.12126032263040543,
+      "learning_rate": 4.951251239389948e-05,
+      "loss": 0.4994,
+      "num_input_tokens_seen": 2795664,
+      "step": 68
+    },
+    {
+      "epoch": 1.2568807339449541,
+      "grad_norm": 0.1069057360291481,
+      "learning_rate": 4.949811761552074e-05,
+      "loss": 0.3275,
+      "num_input_tokens_seen": 2840936,
+      "step": 69
+    },
+    {
+      "epoch": 1.2752293577981653,
+      "grad_norm": 0.10893313586711884,
+      "learning_rate": 4.948351554413879e-05,
+      "loss": 0.4366,
+      "num_input_tokens_seen": 2886768,
+      "step": 70
+    },
+    {
+      "epoch": 1.2935779816513762,
+      "grad_norm": 0.12898756563663483,
+      "learning_rate": 4.9468706303310355e-05,
+      "loss": 0.3916,
+      "num_input_tokens_seen": 2919328,
+      "step": 71
+    },
+    {
+      "epoch": 1.311926605504587,
+      "grad_norm": 0.12405356019735336,
+      "learning_rate": 4.9453690018345144e-05,
+      "loss": 0.3249,
+      "num_input_tokens_seen": 2966744,
+      "step": 72
+    },
+    {
+      "epoch": 1.3302752293577982,
+      "grad_norm": 0.13137595355510712,
+      "learning_rate": 4.943846681630479e-05,
+      "loss": 0.3956,
+      "num_input_tokens_seen": 3007248,
+      "step": 73
+    },
+    {
+      "epoch": 1.3486238532110093,
+      "grad_norm": 0.13920250535011292,
+      "learning_rate": 4.942303682600178e-05,
+      "loss": 0.3956,
+      "num_input_tokens_seen": 3050960,
+      "step": 74
+    },
+    {
+      "epoch": 1.3669724770642202,
+      "grad_norm": 0.1255589872598648,
+      "learning_rate": 4.940740017799833e-05,
+      "loss": 0.3773,
+      "num_input_tokens_seen": 3088592,
+      "step": 75
+    },
+    {
+      "epoch": 1.385321100917431,
+      "grad_norm": 0.10222747176885605,
+      "learning_rate": 4.939155700460536e-05,
+      "loss": 0.4,
+      "num_input_tokens_seen": 3153520,
+      "step": 76
+    },
+    {
+      "epoch": 1.4036697247706422,
+      "grad_norm": 0.13205283880233765,
+      "learning_rate": 4.9375507439881266e-05,
+      "loss": 0.4343,
+      "num_input_tokens_seen": 3199272,
+      "step": 77
+    },
+    {
+      "epoch": 1.4220183486238533,
+      "grad_norm": 0.11005694419145584,
+      "learning_rate": 4.9359251619630886e-05,
+      "loss": 0.3881,
+      "num_input_tokens_seen": 3247128,
+      "step": 78
+    },
+    {
+      "epoch": 1.4403669724770642,
+      "grad_norm": 0.14799247682094574,
+      "learning_rate": 4.9342789681404275e-05,
+      "loss": 0.3972,
+      "num_input_tokens_seen": 3294192,
+      "step": 79
+    },
+    {
+      "epoch": 1.4587155963302751,
+      "grad_norm": 0.1279418021440506,
+      "learning_rate": 4.9326121764495596e-05,
+      "loss": 0.3438,
+      "num_input_tokens_seen": 3329736,
+      "step": 80
+    },
+    {
+      "epoch": 1.4770642201834863,
+      "grad_norm": 0.11036807298660278,
+      "learning_rate": 4.9309248009941914e-05,
+      "loss": 0.3189,
+      "num_input_tokens_seen": 3371376,
+      "step": 81
+    },
+    {
+      "epoch": 1.4954128440366974,
+      "grad_norm": 0.11855707317590714,
+      "learning_rate": 4.9292168560522014e-05,
+      "loss": 0.401,
+      "num_input_tokens_seen": 3412368,
+      "step": 82
+    },
+    {
+      "epoch": 1.5137614678899083,
+      "grad_norm": 0.13195356726646423,
+      "learning_rate": 4.9274883560755156e-05,
+      "loss": 0.4973,
+      "num_input_tokens_seen": 3455000,
+      "step": 83
+    },
+    {
+      "epoch": 1.5321100917431192,
+      "grad_norm": 0.1462787538766861,
+      "learning_rate": 4.925739315689991e-05,
+      "loss": 0.3768,
+      "num_input_tokens_seen": 3488960,
+      "step": 84
+    },
+    {
+      "epoch": 1.5504587155963303,
+      "grad_norm": 0.13765974342823029,
+      "learning_rate": 4.92396974969529e-05,
+      "loss": 0.2999,
+      "num_input_tokens_seen": 3521320,
+      "step": 85
+    },
+    {
+      "epoch": 1.5688073394495414,
+      "grad_norm": 0.15276113152503967,
+      "learning_rate": 4.9221796730647516e-05,
+      "loss": 0.3638,
+      "num_input_tokens_seen": 3559464,
+      "step": 86
+    },
+    {
+      "epoch": 1.5871559633027523,
+      "grad_norm": 0.1441674381494522,
+      "learning_rate": 4.92036910094527e-05,
+      "loss": 0.3919,
+      "num_input_tokens_seen": 3598080,
+      "step": 87
+    },
+    {
+      "epoch": 1.6055045871559632,
+      "grad_norm": 0.1780252456665039,
+      "learning_rate": 4.9185380486571595e-05,
+      "loss": 0.3626,
+      "num_input_tokens_seen": 3630064,
+      "step": 88
+    },
+    {
+      "epoch": 1.6238532110091743,
+      "grad_norm": 0.16947726905345917,
+      "learning_rate": 4.916686531694035e-05,
+      "loss": 0.3439,
+      "num_input_tokens_seen": 3661408,
+      "step": 89
+    },
+    {
+      "epoch": 1.6422018348623855,
+      "grad_norm": 0.1552971601486206,
+      "learning_rate": 4.914814565722671e-05,
+      "loss": 0.3236,
+      "num_input_tokens_seen": 3695480,
+      "step": 90
+    },
+    {
+      "epoch": 1.6605504587155964,
+      "grad_norm": 0.14925938844680786,
+      "learning_rate": 4.912922166582874e-05,
+      "loss": 0.4255,
+      "num_input_tokens_seen": 3734560,
+      "step": 91
+    },
+    {
+      "epoch": 1.6788990825688073,
+      "grad_norm": 0.1332874596118927,
+      "learning_rate": 4.9110093502873476e-05,
+      "loss": 0.3061,
+      "num_input_tokens_seen": 3773112,
+      "step": 92
+    },
+    {
+      "epoch": 1.6972477064220184,
+      "grad_norm": 0.15471243858337402,
+      "learning_rate": 4.909076133021557e-05,
+      "loss": 0.3275,
+      "num_input_tokens_seen": 3813392,
+      "step": 93
+    },
+    {
+      "epoch": 1.7155963302752295,
+      "grad_norm": 0.16010524332523346,
+      "learning_rate": 4.907122531143594e-05,
+      "loss": 0.4179,
+      "num_input_tokens_seen": 3856416,
+      "step": 94
+    },
+    {
+      "epoch": 1.7339449541284404,
+      "grad_norm": 0.13423003256320953,
+      "learning_rate": 4.905148561184033e-05,
+      "loss": 0.3593,
+      "num_input_tokens_seen": 3899472,
+      "step": 95
+    },
+    {
+      "epoch": 1.7522935779816513,
+      "grad_norm": 0.14900773763656616,
+      "learning_rate": 4.9031542398457974e-05,
+      "loss": 0.5007,
+      "num_input_tokens_seen": 3962976,
+      "step": 96
+    },
+    {
+      "epoch": 1.7706422018348624,
+      "grad_norm": 0.15728624165058136,
+      "learning_rate": 4.9011395840040144e-05,
+      "loss": 0.3484,
+      "num_input_tokens_seen": 4000696,
+      "step": 97
+    },
+    {
+      "epoch": 1.7889908256880735,
+      "grad_norm": 0.11092367768287659,
+      "learning_rate": 4.8991046107058735e-05,
+      "loss": 0.2889,
+      "num_input_tokens_seen": 4045256,
+      "step": 98
+    },
+    {
+      "epoch": 1.8073394495412844,
+      "grad_norm": 0.1289113610982895,
+      "learning_rate": 4.8970493371704826e-05,
+      "loss": 0.2203,
+      "num_input_tokens_seen": 4076800,
+      "step": 99
+    },
+    {
+      "epoch": 1.8256880733944953,
+      "grad_norm": 0.18886639177799225,
+      "learning_rate": 4.894973780788722e-05,
+      "loss": 0.3966,
+      "num_input_tokens_seen": 4119840,
+      "step": 100
+    },
+    {
+      "epoch": 1.8440366972477065,
+      "grad_norm": 0.1563039869070053,
+      "learning_rate": 4.892877959123097e-05,
+      "loss": 0.4417,
+      "num_input_tokens_seen": 4165848,
+      "step": 101
+    },
+    {
+      "epoch": 1.8623853211009176,
+      "grad_norm": 0.16883380711078644,
+      "learning_rate": 4.890761889907589e-05,
+      "loss": 0.4258,
+      "num_input_tokens_seen": 4202824,
+      "step": 102
+    },
+    {
+      "epoch": 1.8807339449541285,
+      "grad_norm": 0.18241995573043823,
+      "learning_rate": 4.8886255910475054e-05,
+      "loss": 0.3952,
+      "num_input_tokens_seen": 4233888,
+      "step": 103
+    },
+    {
+      "epoch": 1.8990825688073394,
+      "grad_norm": 0.19913265109062195,
+      "learning_rate": 4.88646908061933e-05,
+      "loss": 0.3241,
+      "num_input_tokens_seen": 4267064,
+      "step": 104
+    },
+    {
+      "epoch": 1.9174311926605505,
+      "grad_norm": 0.18295545876026154,
+      "learning_rate": 4.884292376870567e-05,
+      "loss": 0.4239,
+      "num_input_tokens_seen": 4312536,
+      "step": 105
+    },
+    {
+      "epoch": 1.9357798165137616,
+      "grad_norm": 0.16657495498657227,
+      "learning_rate": 4.8820954982195905e-05,
+      "loss": 0.2579,
+      "num_input_tokens_seen": 4356656,
+      "step": 106
+    },
+    {
+      "epoch": 1.9541284403669725,
+      "grad_norm": 0.18504932522773743,
+      "learning_rate": 4.879878463255483e-05,
+      "loss": 0.44,
+      "num_input_tokens_seen": 4400216,
+      "step": 107
+    },
+    {
+      "epoch": 1.9724770642201834,
+      "grad_norm": 0.1923118382692337,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.2662,
+      "num_input_tokens_seen": 4436968,
+      "step": 108
+    },
+    {
+      "epoch": 1.9908256880733946,
+      "grad_norm": 0.19636788964271545,
+      "learning_rate": 4.875383999596828e-05,
+      "loss": 0.4211,
+      "num_input_tokens_seen": 4488232,
+      "step": 109
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3168099820613861,
+      "learning_rate": 4.873106608932585e-05,
+      "loss": 0.2499,
+      "num_input_tokens_seen": 4518176,
+      "step": 110
+    },
+    {
+      "epoch": 2.018348623853211,
+      "grad_norm": 0.14410308003425598,
+      "learning_rate": 4.8708091380154984e-05,
+      "loss": 0.2722,
+      "num_input_tokens_seen": 4570896,
+      "step": 111
+    },
+    {
+      "epoch": 2.036697247706422,
+      "grad_norm": 0.17840909957885742,
+      "learning_rate": 4.868491606285823e-05,
+      "loss": 0.2758,
+      "num_input_tokens_seen": 4613576,
+      "step": 112
+    },
+    {
+      "epoch": 2.055045871559633,
+      "grad_norm": 0.178523987531662,
+      "learning_rate": 4.866154033353561e-05,
+      "loss": 0.3361,
+      "num_input_tokens_seen": 4652896,
+      "step": 113
+    },
+    {
+      "epoch": 2.073394495412844,
+      "grad_norm": 0.17396725714206696,
+      "learning_rate": 4.8637964389982926e-05,
+      "loss": 0.2667,
+      "num_input_tokens_seen": 4694256,
+      "step": 114
+    },
+    {
+      "epoch": 2.091743119266055,
+      "grad_norm": 0.19471587240695953,
+      "learning_rate": 4.8614188431690125e-05,
+      "loss": 0.3628,
+      "num_input_tokens_seen": 4747552,
+      "step": 115
+    },
+    {
+      "epoch": 2.1100917431192663,
+      "grad_norm": 0.1722450852394104,
+      "learning_rate": 4.859021265983959e-05,
+      "loss": 0.3599,
+      "num_input_tokens_seen": 4794080,
+      "step": 116
+    },
+    {
+      "epoch": 2.128440366972477,
+      "grad_norm": 0.20137006044387817,
+      "learning_rate": 4.856603727730447e-05,
+      "loss": 0.4262,
+      "num_input_tokens_seen": 4847912,
+      "step": 117
+    },
+    {
+      "epoch": 2.146788990825688,
+      "grad_norm": 0.19395771622657776,
+      "learning_rate": 4.854166248864689e-05,
+      "loss": 0.3118,
+      "num_input_tokens_seen": 4885480,
+      "step": 118
+    },
+    {
+      "epoch": 2.165137614678899,
+      "grad_norm": 0.209548681974411,
+      "learning_rate": 4.85170885001163e-05,
+      "loss": 0.3725,
+      "num_input_tokens_seen": 4921240,
+      "step": 119
+    },
+    {
+      "epoch": 2.18348623853211,
+      "grad_norm": 0.18228279054164886,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.3816,
+      "num_input_tokens_seen": 4960224,
+      "step": 120
+    },
+    {
+      "epoch": 2.2018348623853212,
+      "grad_norm": 0.24349354207515717,
+      "learning_rate": 4.846734375685989e-05,
+      "loss": 0.3383,
+      "num_input_tokens_seen": 4990536,
+      "step": 121
+    },
+    {
+      "epoch": 2.220183486238532,
+      "grad_norm": 0.17600344121456146,
+      "learning_rate": 4.844217342305363e-05,
+      "loss": 0.3011,
+      "num_input_tokens_seen": 5044296,
+      "step": 122
+    },
+    {
+      "epoch": 2.238532110091743,
+      "grad_norm": 0.18766675889492035,
+      "learning_rate": 4.8416804731209945e-05,
+      "loss": 0.4458,
+      "num_input_tokens_seen": 5088368,
+      "step": 123
+    },
+    {
+      "epoch": 2.2568807339449544,
+      "grad_norm": 0.17657820880413055,
+      "learning_rate": 4.839123789598829e-05,
+      "loss": 0.2564,
+      "num_input_tokens_seen": 5133472,
+      "step": 124
+    },
+    {
+      "epoch": 2.2752293577981653,
+      "grad_norm": 0.20606014132499695,
+      "learning_rate": 4.836547313372471e-05,
+      "loss": 0.313,
+      "num_input_tokens_seen": 5167768,
+      "step": 125
+    },
+    {
+      "epoch": 2.293577981651376,
+      "grad_norm": 0.23511061072349548,
+      "learning_rate": 4.8339510662430046e-05,
+      "loss": 0.2963,
+      "num_input_tokens_seen": 5209400,
+      "step": 126
+    },
+    {
+      "epoch": 2.311926605504587,
+      "grad_norm": 0.18234293162822723,
+      "learning_rate": 4.8313350701788054e-05,
+      "loss": 0.2566,
+      "num_input_tokens_seen": 5249360,
+      "step": 127
+    },
+    {
+      "epoch": 2.330275229357798,
+      "grad_norm": 0.2223992496728897,
+      "learning_rate": 4.828699347315356e-05,
+      "loss": 0.2833,
+      "num_input_tokens_seen": 5300808,
+      "step": 128
+    },
+    {
+      "epoch": 2.3486238532110093,
+      "grad_norm": 0.23101739585399628,
+      "learning_rate": 4.826043919955062e-05,
+      "loss": 0.3099,
+      "num_input_tokens_seen": 5332960,
+      "step": 129
+    },
+    {
+      "epoch": 2.36697247706422,
+      "grad_norm": 0.26640889048576355,
+      "learning_rate": 4.823368810567056e-05,
+      "loss": 0.3238,
+      "num_input_tokens_seen": 5365008,
+      "step": 130
+    },
+    {
+      "epoch": 2.385321100917431,
+      "grad_norm": 0.2374572902917862,
+      "learning_rate": 4.820674041787017e-05,
+      "loss": 0.3153,
+      "num_input_tokens_seen": 5400184,
+      "step": 131
+    },
+    {
+      "epoch": 2.4036697247706424,
+      "grad_norm": 0.22812288999557495,
+      "learning_rate": 4.817959636416969e-05,
+      "loss": 0.2997,
+      "num_input_tokens_seen": 5440320,
+      "step": 132
+    },
+    {
+      "epoch": 2.4220183486238533,
+      "grad_norm": 0.20079149305820465,
+      "learning_rate": 4.815225617425095e-05,
+      "loss": 0.2373,
+      "num_input_tokens_seen": 5480832,
+      "step": 133
+    },
+    {
+      "epoch": 2.4403669724770642,
+      "grad_norm": 0.196709543466568,
+      "learning_rate": 4.81247200794554e-05,
+      "loss": 0.2456,
+      "num_input_tokens_seen": 5526936,
+      "step": 134
+    },
+    {
+      "epoch": 2.458715596330275,
+      "grad_norm": 0.17305873334407806,
+      "learning_rate": 4.8096988312782174e-05,
+      "loss": 0.2099,
+      "num_input_tokens_seen": 5566384,
+      "step": 135
+    },
+    {
+      "epoch": 2.477064220183486,
+      "grad_norm": 3.584635019302368,
+      "learning_rate": 4.806906110888606e-05,
+      "loss": 0.3485,
+      "num_input_tokens_seen": 5629896,
+      "step": 136
+    },
+    {
+      "epoch": 2.4954128440366974,
+      "grad_norm": 0.23481500148773193,
+      "learning_rate": 4.80409387040756e-05,
+      "loss": 0.2231,
+      "num_input_tokens_seen": 5674504,
+      "step": 137
+    },
+    {
+      "epoch": 2.5137614678899083,
+      "grad_norm": 0.27899855375289917,
+      "learning_rate": 4.8012621336311016e-05,
+      "loss": 0.4285,
+      "num_input_tokens_seen": 5714000,
+      "step": 138
+    },
+    {
+      "epoch": 2.532110091743119,
+      "grad_norm": 0.24404938519001007,
+      "learning_rate": 4.798410924520223e-05,
+      "loss": 0.3343,
+      "num_input_tokens_seen": 5756856,
+      "step": 139
+    },
+    {
+      "epoch": 2.5504587155963305,
+      "grad_norm": 0.26869162917137146,
+      "learning_rate": 4.7955402672006854e-05,
+      "loss": 0.2497,
+      "num_input_tokens_seen": 5781192,
+      "step": 140
+    },
+    {
+      "epoch": 2.5688073394495414,
+      "grad_norm": 0.2057972550392151,
+      "learning_rate": 4.79265018596281e-05,
+      "loss": 0.2991,
+      "num_input_tokens_seen": 5824024,
+      "step": 141
+    },
+    {
+      "epoch": 2.5871559633027523,
+      "grad_norm": 0.2184937596321106,
+      "learning_rate": 4.789740705261278e-05,
+      "loss": 0.2406,
+      "num_input_tokens_seen": 5862584,
+      "step": 142
+    },
+    {
+      "epoch": 2.6055045871559632,
+      "grad_norm": 0.23603741824626923,
+      "learning_rate": 4.786811849714918e-05,
+      "loss": 0.2722,
+      "num_input_tokens_seen": 5897344,
+      "step": 143
+    },
+    {
+      "epoch": 2.623853211009174,
+      "grad_norm": 0.22983981668949127,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 0.374,
+      "num_input_tokens_seen": 5931736,
+      "step": 144
+    },
+    {
+      "epoch": 2.6422018348623855,
+      "grad_norm": 0.2825419306755066,
+      "learning_rate": 4.780896113382536e-05,
+      "loss": 0.3386,
+      "num_input_tokens_seen": 5972784,
+      "step": 145
+    },
+    {
+      "epoch": 2.6605504587155964,
+      "grad_norm": 0.4502134621143341,
+      "learning_rate": 4.777909282653042e-05,
+      "loss": 0.2289,
+      "num_input_tokens_seen": 6018968,
+      "step": 146
+    },
+    {
+      "epoch": 2.6788990825688073,
+      "grad_norm": 0.2428288459777832,
+      "learning_rate": 4.7749031771913584e-05,
+      "loss": 0.4061,
+      "num_input_tokens_seen": 6062520,
+      "step": 147
+    },
+    {
+      "epoch": 2.6972477064220186,
+      "grad_norm": 0.2685629725456238,
+      "learning_rate": 4.771877822433911e-05,
+      "loss": 0.2198,
+      "num_input_tokens_seen": 6087928,
+      "step": 148
+    },
+    {
+      "epoch": 2.7155963302752295,
+      "grad_norm": 0.24021446704864502,
+      "learning_rate": 4.7688332439800096e-05,
+      "loss": 0.34,
+      "num_input_tokens_seen": 6134792,
+      "step": 149
+    },
+    {
+      "epoch": 2.7339449541284404,
+      "grad_norm": 0.2568534314632416,
+      "learning_rate": 4.765769467591625e-05,
+      "loss": 0.3292,
+      "num_input_tokens_seen": 6183296,
+      "step": 150
+    },
+    {
+      "epoch": 2.7522935779816513,
+      "grad_norm": 0.20823974907398224,
+      "learning_rate": 4.762686519193175e-05,
+      "loss": 0.2539,
+      "num_input_tokens_seen": 6225840,
+      "step": 151
+    },
+    {
+      "epoch": 2.770642201834862,
+      "grad_norm": 0.23333317041397095,
+      "learning_rate": 4.759584424871302e-05,
+      "loss": 0.3571,
+      "num_input_tokens_seen": 6274760,
+      "step": 152
+    },
+    {
+      "epoch": 2.7889908256880735,
+      "grad_norm": 0.20232398808002472,
+      "learning_rate": 4.756463210874652e-05,
+      "loss": 0.2783,
+      "num_input_tokens_seen": 6326168,
+      "step": 153
+    },
+    {
+      "epoch": 2.8073394495412844,
+      "grad_norm": 0.3479433059692383,
+      "learning_rate": 4.7533229036136553e-05,
+      "loss": 0.2925,
+      "num_input_tokens_seen": 6360312,
+      "step": 154
+    },
+    {
+      "epoch": 2.8256880733944953,
+      "grad_norm": 0.2659524083137512,
+      "learning_rate": 4.750163529660303e-05,
+      "loss": 0.2606,
+      "num_input_tokens_seen": 6395496,
+      "step": 155
+    },
+    {
+      "epoch": 2.8440366972477067,
+      "grad_norm": 0.24823158979415894,
+      "learning_rate": 4.7469851157479177e-05,
+      "loss": 0.3721,
+      "num_input_tokens_seen": 6437064,
+      "step": 156
+    },
+    {
+      "epoch": 2.8623853211009176,
+      "grad_norm": 0.32034072279930115,
+      "learning_rate": 4.743787688770932e-05,
+      "loss": 0.3931,
+      "num_input_tokens_seen": 6477616,
+      "step": 157
+    },
+    {
+      "epoch": 2.8807339449541285,
+      "grad_norm": 0.23295725882053375,
+      "learning_rate": 4.740571275784659e-05,
+      "loss": 0.2201,
+      "num_input_tokens_seen": 6518680,
+      "step": 158
+    },
+    {
+      "epoch": 2.8990825688073394,
+      "grad_norm": 0.2758423984050751,
+      "learning_rate": 4.737335904005063e-05,
+      "loss": 0.2579,
+      "num_input_tokens_seen": 6549768,
+      "step": 159
+    },
+    {
+      "epoch": 2.9174311926605503,
+      "grad_norm": 0.26690953969955444,
+      "learning_rate": 4.734081600808531e-05,
+      "loss": 0.2575,
+      "num_input_tokens_seen": 6581000,
+      "step": 160
+    },
+    {
+      "epoch": 2.9357798165137616,
+      "grad_norm": 0.26657482981681824,
+      "learning_rate": 4.730808393731639e-05,
+      "loss": 0.2597,
+      "num_input_tokens_seen": 6612632,
+      "step": 161
+    },
+    {
+      "epoch": 2.9541284403669725,
+      "grad_norm": 0.22647295892238617,
+      "learning_rate": 4.72751631047092e-05,
+      "loss": 0.3335,
+      "num_input_tokens_seen": 6654288,
+      "step": 162
+    },
+    {
+      "epoch": 2.9724770642201834,
+      "grad_norm": 0.2863366901874542,
+      "learning_rate": 4.72420537888263e-05,
+      "loss": 0.374,
+      "num_input_tokens_seen": 6707208,
+      "step": 163
+    },
+    {
+      "epoch": 2.9908256880733948,
+      "grad_norm": 0.2606408894062042,
+      "learning_rate": 4.7208756269825104e-05,
+      "loss": 0.3477,
+      "num_input_tokens_seen": 6748448,
+      "step": 164
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.440924733877182,
+      "learning_rate": 4.717527082945554e-05,
+      "loss": 0.3214,
+      "num_input_tokens_seen": 6777264,
+      "step": 165
+    },
+    {
+      "epoch": 3.018348623853211,
+      "grad_norm": 0.27583903074264526,
+      "learning_rate": 4.714159775105765e-05,
+      "loss": 0.2681,
+      "num_input_tokens_seen": 6809456,
+      "step": 166
+    },
+    {
+      "epoch": 3.036697247706422,
+      "grad_norm": 0.2995987832546234,
+      "learning_rate": 4.7107737319559176e-05,
+      "loss": 0.2633,
+      "num_input_tokens_seen": 6845768,
+      "step": 167
+    },
+    {
+      "epoch": 3.055045871559633,
+      "grad_norm": 0.23999951779842377,
+      "learning_rate": 4.707368982147318e-05,
+      "loss": 0.1961,
+      "num_input_tokens_seen": 6893056,
+      "step": 168
+    },
+    {
+      "epoch": 3.073394495412844,
+      "grad_norm": 0.23356525599956512,
+      "learning_rate": 4.703945554489558e-05,
+      "loss": 0.2836,
+      "num_input_tokens_seen": 6932480,
+      "step": 169
+    },
+    {
+      "epoch": 3.091743119266055,
+      "grad_norm": 0.29919493198394775,
+      "learning_rate": 4.700503477950278e-05,
+      "loss": 0.2838,
+      "num_input_tokens_seen": 6975992,
+      "step": 170
+    },
+    {
+      "epoch": 3.1100917431192663,
+      "grad_norm": 0.3350690007209778,
+      "learning_rate": 4.697042781654913e-05,
+      "loss": 0.3489,
+      "num_input_tokens_seen": 7021840,
+      "step": 171
+    },
+    {
+      "epoch": 3.128440366972477,
+      "grad_norm": 0.2837466895580292,
+      "learning_rate": 4.693563494886455e-05,
+      "loss": 0.3797,
+      "num_input_tokens_seen": 7065192,
+      "step": 172
+    },
+    {
+      "epoch": 3.146788990825688,
+      "grad_norm": 0.24601787328720093,
+      "learning_rate": 4.6900656470851964e-05,
+      "loss": 0.2046,
+      "num_input_tokens_seen": 7114544,
+      "step": 173
+    },
+    {
+      "epoch": 3.165137614678899,
+      "grad_norm": 0.32290250062942505,
+      "learning_rate": 4.6865492678484895e-05,
+      "loss": 0.2596,
+      "num_input_tokens_seen": 7152736,
+      "step": 174
+    },
+    {
+      "epoch": 3.18348623853211,
+      "grad_norm": 0.33591920137405396,
+      "learning_rate": 4.68301438693049e-05,
+      "loss": 0.3045,
+      "num_input_tokens_seen": 7207464,
+      "step": 175
+    },
+    {
+      "epoch": 3.2018348623853212,
+      "grad_norm": 0.25471043586730957,
+      "learning_rate": 4.679461034241906e-05,
+      "loss": 0.2096,
+      "num_input_tokens_seen": 7238640,
+      "step": 176
+    },
+    {
+      "epoch": 3.220183486238532,
+      "grad_norm": 0.31238994002342224,
+      "learning_rate": 4.6758892398497494e-05,
+      "loss": 0.2226,
+      "num_input_tokens_seen": 7279112,
+      "step": 177
+    },
+    {
+      "epoch": 3.238532110091743,
+      "grad_norm": 0.35679712891578674,
+      "learning_rate": 4.672299033977076e-05,
+      "loss": 0.2403,
+      "num_input_tokens_seen": 7311632,
+      "step": 178
+    },
+    {
+      "epoch": 3.2568807339449544,
+      "grad_norm": 0.326914519071579,
+      "learning_rate": 4.6686904470027316e-05,
+      "loss": 0.2156,
+      "num_input_tokens_seen": 7344864,
+      "step": 179
+    },
+    {
+      "epoch": 3.2752293577981653,
+      "grad_norm": 0.3293381929397583,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.238,
+      "num_input_tokens_seen": 7389944,
+      "step": 180
+    },
+    {
+      "epoch": 3.293577981651376,
+      "grad_norm": 0.3313307762145996,
+      "learning_rate": 4.661418252041827e-05,
+      "loss": 0.2251,
+      "num_input_tokens_seen": 7423672,
+      "step": 181
+    },
+    {
+      "epoch": 3.311926605504587,
+      "grad_norm": 0.3328595459461212,
+      "learning_rate": 4.657754705589591e-05,
+      "loss": 0.2922,
+      "num_input_tokens_seen": 7459576,
+      "step": 182
+    },
+    {
+      "epoch": 3.330275229357798,
+      "grad_norm": 0.2721710801124573,
+      "learning_rate": 4.6540729011038146e-05,
+      "loss": 0.2698,
+      "num_input_tokens_seen": 7511736,
+      "step": 183
+    },
+    {
+      "epoch": 3.3486238532110093,
+      "grad_norm": 0.2488890290260315,
+      "learning_rate": 4.650372869738414e-05,
+      "loss": 0.173,
+      "num_input_tokens_seen": 7558552,
+      "step": 184
+    },
+    {
+      "epoch": 3.36697247706422,
+      "grad_norm": 0.3800615668296814,
+      "learning_rate": 4.6466546428015336e-05,
+      "loss": 0.32,
+      "num_input_tokens_seen": 7599040,
+      "step": 185
+    },
+    {
+      "epoch": 3.385321100917431,
+      "grad_norm": 0.3377014100551605,
+      "learning_rate": 4.642918251755281e-05,
+      "loss": 0.3058,
+      "num_input_tokens_seen": 7653264,
+      "step": 186
+    },
+    {
+      "epoch": 3.4036697247706424,
+      "grad_norm": 0.25239789485931396,
+      "learning_rate": 4.639163728215463e-05,
+      "loss": 0.1896,
+      "num_input_tokens_seen": 7694272,
+      "step": 187
+    },
+    {
+      "epoch": 3.4220183486238533,
+      "grad_norm": 0.34607502818107605,
+      "learning_rate": 4.6353911039513145e-05,
+      "loss": 0.2933,
+      "num_input_tokens_seen": 7730848,
+      "step": 188
+    },
+    {
+      "epoch": 3.4403669724770642,
+      "grad_norm": 0.30653324723243713,
+      "learning_rate": 4.6316004108852305e-05,
+      "loss": 0.2625,
+      "num_input_tokens_seen": 7781200,
+      "step": 189
+    },
+    {
+      "epoch": 3.458715596330275,
+      "grad_norm": 0.2943236231803894,
+      "learning_rate": 4.627791681092499e-05,
+      "loss": 0.3372,
+      "num_input_tokens_seen": 7825032,
+      "step": 190
+    },
+    {
+      "epoch": 3.477064220183486,
+      "grad_norm": 0.30080685019493103,
+      "learning_rate": 4.623964946801027e-05,
+      "loss": 0.2229,
+      "num_input_tokens_seen": 7855840,
+      "step": 191
+    },
+    {
+      "epoch": 3.4954128440366974,
+      "grad_norm": 0.3511403799057007,
+      "learning_rate": 4.620120240391065e-05,
+      "loss": 0.3967,
+      "num_input_tokens_seen": 7905928,
+      "step": 192
+    },
+    {
+      "epoch": 3.5137614678899083,
+      "grad_norm": 0.273583322763443,
+      "learning_rate": 4.61625759439494e-05,
+      "loss": 0.2254,
+      "num_input_tokens_seen": 7955992,
+      "step": 193
+    },
+    {
+      "epoch": 3.532110091743119,
+      "grad_norm": 0.3457902669906616,
+      "learning_rate": 4.612377041496776e-05,
+      "loss": 0.2553,
+      "num_input_tokens_seen": 7998024,
+      "step": 194
+    },
+    {
+      "epoch": 3.5504587155963305,
+      "grad_norm": 0.31968954205513,
+      "learning_rate": 4.608478614532215e-05,
+      "loss": 0.2197,
+      "num_input_tokens_seen": 8055672,
+      "step": 195
+    },
+    {
+      "epoch": 3.5688073394495414,
+      "grad_norm": 0.34753403067588806,
+      "learning_rate": 4.604562346488144e-05,
+      "loss": 0.2507,
+      "num_input_tokens_seen": 8090848,
+      "step": 196
+    },
+    {
+      "epoch": 3.5871559633027523,
+      "grad_norm": 0.3808669149875641,
+      "learning_rate": 4.6006282705024144e-05,
+      "loss": 0.2422,
+      "num_input_tokens_seen": 8136680,
+      "step": 197
+    },
+    {
+      "epoch": 3.6055045871559632,
+      "grad_norm": 0.3004499673843384,
+      "learning_rate": 4.5966764198635606e-05,
+      "loss": 0.2107,
+      "num_input_tokens_seen": 8187472,
+      "step": 198
+    },
+    {
+      "epoch": 3.623853211009174,
+      "grad_norm": 0.30718186497688293,
+      "learning_rate": 4.592706828010518e-05,
+      "loss": 0.1854,
+      "num_input_tokens_seen": 8225216,
+      "step": 199
+    },
+    {
+      "epoch": 3.6422018348623855,
+      "grad_norm": 0.23112858831882477,
+      "learning_rate": 4.588719528532342e-05,
+      "loss": 0.1687,
+      "num_input_tokens_seen": 8274456,
+      "step": 200
+    },
+    {
+      "epoch": 3.6605504587155964,
+      "grad_norm": 0.25966888666152954,
+      "learning_rate": 4.5847145551679206e-05,
+      "loss": 0.2549,
+      "num_input_tokens_seen": 8317016,
+      "step": 201
+    },
+    {
+      "epoch": 3.6788990825688073,
+      "grad_norm": 0.25600987672805786,
+      "learning_rate": 4.580691941805695e-05,
+      "loss": 0.1602,
+      "num_input_tokens_seen": 8361856,
+      "step": 202
+    },
+    {
+      "epoch": 3.6972477064220186,
+      "grad_norm": 0.33986184000968933,
+      "learning_rate": 4.5766517224833637e-05,
+      "loss": 0.2495,
+      "num_input_tokens_seen": 8410696,
+      "step": 203
+    },
+    {
+      "epoch": 3.7155963302752295,
+      "grad_norm": 0.36899781227111816,
+      "learning_rate": 4.572593931387604e-05,
+      "loss": 0.2012,
+      "num_input_tokens_seen": 8441872,
+      "step": 204
+    },
+    {
+      "epoch": 3.7339449541284404,
+      "grad_norm": 0.42072632908821106,
+      "learning_rate": 4.568518602853776e-05,
+      "loss": 0.2373,
+      "num_input_tokens_seen": 8482544,
+      "step": 205
+    },
+    {
+      "epoch": 3.7522935779816513,
+      "grad_norm": 0.40593233704566956,
+      "learning_rate": 4.5644257713656356e-05,
+      "loss": 0.233,
+      "num_input_tokens_seen": 8519856,
+      "step": 206
+    },
+    {
+      "epoch": 3.770642201834862,
+      "grad_norm": 0.38003161549568176,
+      "learning_rate": 4.5603154715550386e-05,
+      "loss": 0.225,
+      "num_input_tokens_seen": 8551392,
+      "step": 207
+    },
+    {
+      "epoch": 3.7889908256880735,
+      "grad_norm": 0.2564244568347931,
+      "learning_rate": 4.556187738201656e-05,
+      "loss": 0.2975,
+      "num_input_tokens_seen": 8599472,
+      "step": 208
+    },
+    {
+      "epoch": 3.8073394495412844,
+      "grad_norm": 0.29023391008377075,
+      "learning_rate": 4.552042606232668e-05,
+      "loss": 0.2033,
+      "num_input_tokens_seen": 8631880,
+      "step": 209
+    },
+    {
+      "epoch": 3.8256880733944953,
+      "grad_norm": 0.32886001467704773,
+      "learning_rate": 4.54788011072248e-05,
+      "loss": 0.2024,
+      "num_input_tokens_seen": 8675016,
+      "step": 210
+    },
+    {
+      "epoch": 3.8440366972477067,
+      "grad_norm": 0.36749884486198425,
+      "learning_rate": 4.5437002868924166e-05,
+      "loss": 0.2304,
+      "num_input_tokens_seen": 8713248,
+      "step": 211
+    },
+    {
+      "epoch": 3.8623853211009176,
+      "grad_norm": 0.3055097758769989,
+      "learning_rate": 4.539503170110431e-05,
+      "loss": 0.2928,
+      "num_input_tokens_seen": 8748800,
+      "step": 212
+    },
+    {
+      "epoch": 3.8807339449541285,
+      "grad_norm": 0.38436686992645264,
+      "learning_rate": 4.535288795890798e-05,
+      "loss": 0.2214,
+      "num_input_tokens_seen": 8787832,
+      "step": 213
+    },
+    {
+      "epoch": 3.8990825688073394,
+      "grad_norm": 0.44330883026123047,
+      "learning_rate": 4.531057199893824e-05,
+      "loss": 0.2168,
+      "num_input_tokens_seen": 8819616,
+      "step": 214
+    },
+    {
+      "epoch": 3.9174311926605503,
+      "grad_norm": 0.28318527340888977,
+      "learning_rate": 4.526808417925531e-05,
+      "loss": 0.279,
+      "num_input_tokens_seen": 8860744,
+      "step": 215
+    },
+    {
+      "epoch": 3.9357798165137616,
+      "grad_norm": 0.3287319839000702,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.2597,
+      "num_input_tokens_seen": 8906432,
+      "step": 216
+    },
+    {
+      "epoch": 3.9541284403669725,
+      "grad_norm": 0.35815751552581787,
+      "learning_rate": 4.5182594400259e-05,
+      "loss": 0.241,
+      "num_input_tokens_seen": 8955104,
+      "step": 217
+    },
+    {
+      "epoch": 3.9724770642201834,
+      "grad_norm": 0.3299608528614044,
+      "learning_rate": 4.5139593164324986e-05,
+      "loss": 0.2157,
+      "num_input_tokens_seen": 8990200,
+      "step": 218
+    },
+    {
+      "epoch": 3.9908256880733948,
+      "grad_norm": 0.2916093170642853,
+      "learning_rate": 4.509642151543043e-05,
+      "loss": 0.2046,
+      "num_input_tokens_seen": 9020760,
+      "step": 219
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.40076637268066406,
+      "learning_rate": 4.50530798188761e-05,
+      "loss": 0.1714,
+      "num_input_tokens_seen": 9036352,
+      "step": 220
+    },
+    {
+      "epoch": 4.018348623853211,
+      "grad_norm": 0.40515249967575073,
+      "learning_rate": 4.50095684414016e-05,
+      "loss": 0.1811,
+      "num_input_tokens_seen": 9091776,
+      "step": 221
+    },
+    {
+      "epoch": 4.036697247706422,
+      "grad_norm": 0.32984718680381775,
+      "learning_rate": 4.496588775118232e-05,
+      "loss": 0.2101,
+      "num_input_tokens_seen": 9134080,
+      "step": 222
+    },
+    {
+      "epoch": 4.055045871559633,
+      "grad_norm": 0.27288541197776794,
+      "learning_rate": 4.4922038117826334e-05,
+      "loss": 0.1444,
+      "num_input_tokens_seen": 9172720,
+      "step": 223
+    },
+    {
+      "epoch": 4.073394495412844,
+      "grad_norm": 0.5168021321296692,
+      "learning_rate": 4.48780199123712e-05,
+      "loss": 0.2342,
+      "num_input_tokens_seen": 9213664,
+      "step": 224
+    },
+    {
+      "epoch": 4.091743119266055,
+      "grad_norm": 0.3986498713493347,
+      "learning_rate": 4.4833833507280884e-05,
+      "loss": 0.1676,
+      "num_input_tokens_seen": 9261768,
+      "step": 225
+    },
+    {
+      "epoch": 4.110091743119266,
+      "grad_norm": 0.4472793936729431,
+      "learning_rate": 4.478947927644258e-05,
+      "loss": 0.295,
+      "num_input_tokens_seen": 9300928,
+      "step": 226
+    },
+    {
+      "epoch": 4.128440366972477,
+      "grad_norm": 0.39240705966949463,
+      "learning_rate": 4.474495759516358e-05,
+      "loss": 0.17,
+      "num_input_tokens_seen": 9329472,
+      "step": 227
+    },
+    {
+      "epoch": 4.146788990825688,
+      "grad_norm": 0.354526549577713,
+      "learning_rate": 4.4700268840168045e-05,
+      "loss": 0.1759,
+      "num_input_tokens_seen": 9365640,
+      "step": 228
+    },
+    {
+      "epoch": 4.165137614678899,
+      "grad_norm": 0.3216766119003296,
+      "learning_rate": 4.4655413389593856e-05,
+      "loss": 0.1878,
+      "num_input_tokens_seen": 9410552,
+      "step": 229
+    },
+    {
+      "epoch": 4.18348623853211,
+      "grad_norm": 0.30976617336273193,
+      "learning_rate": 4.4610391622989396e-05,
+      "loss": 0.1637,
+      "num_input_tokens_seen": 9452416,
+      "step": 230
+    },
+    {
+      "epoch": 4.201834862385321,
+      "grad_norm": 0.385437935590744,
+      "learning_rate": 4.456520392131035e-05,
+      "loss": 0.2748,
+      "num_input_tokens_seen": 9503528,
+      "step": 231
+    },
+    {
+      "epoch": 4.220183486238533,
+      "grad_norm": 0.37948480248451233,
+      "learning_rate": 4.4519850666916484e-05,
+      "loss": 0.2635,
+      "num_input_tokens_seen": 9541592,
+      "step": 232
+    },
+    {
+      "epoch": 4.238532110091743,
+      "grad_norm": 0.36141568422317505,
+      "learning_rate": 4.447433224356839e-05,
+      "loss": 0.2027,
+      "num_input_tokens_seen": 9586064,
+      "step": 233
+    },
+    {
+      "epoch": 4.256880733944954,
+      "grad_norm": 0.4549350440502167,
+      "learning_rate": 4.442864903642428e-05,
+      "loss": 0.2107,
+      "num_input_tokens_seen": 9641688,
+      "step": 234
+    },
+    {
+      "epoch": 4.275229357798165,
+      "grad_norm": 0.3979765474796295,
+      "learning_rate": 4.438280143203665e-05,
+      "loss": 0.2879,
+      "num_input_tokens_seen": 9686240,
+      "step": 235
+    },
+    {
+      "epoch": 4.293577981651376,
+      "grad_norm": 0.35011065006256104,
+      "learning_rate": 4.43367898183491e-05,
+      "loss": 0.2594,
+      "num_input_tokens_seen": 9735632,
+      "step": 236
+    },
+    {
+      "epoch": 4.3119266055045875,
+      "grad_norm": 0.3999224007129669,
+      "learning_rate": 4.4290614584693004e-05,
+      "loss": 0.1907,
+      "num_input_tokens_seen": 9766536,
+      "step": 237
+    },
+    {
+      "epoch": 4.330275229357798,
+      "grad_norm": 0.39629611372947693,
+      "learning_rate": 4.4244276121784195e-05,
+      "loss": 0.1805,
+      "num_input_tokens_seen": 9796400,
+      "step": 238
+    },
+    {
+      "epoch": 4.348623853211009,
+      "grad_norm": 0.36784592270851135,
+      "learning_rate": 4.4197774821719714e-05,
+      "loss": 0.1824,
+      "num_input_tokens_seen": 9831992,
+      "step": 239
+    },
+    {
+      "epoch": 4.36697247706422,
+      "grad_norm": 0.3408430516719818,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.1721,
+      "num_input_tokens_seen": 9875640,
+      "step": 240
+    },
+    {
+      "epoch": 4.385321100917431,
+      "grad_norm": 0.3232553005218506,
+      "learning_rate": 4.410428528539783e-05,
+      "loss": 0.275,
+      "num_input_tokens_seen": 9916816,
+      "step": 241
+    },
+    {
+      "epoch": 4.4036697247706424,
+      "grad_norm": 0.38150206208229065,
+      "learning_rate": 4.405729784021046e-05,
+      "loss": 0.1963,
+      "num_input_tokens_seen": 9962928,
+      "step": 242
+    },
+    {
+      "epoch": 4.422018348623853,
+      "grad_norm": 0.4176963269710541,
+      "learning_rate": 4.401014914000078e-05,
+      "loss": 0.1626,
+      "num_input_tokens_seen": 9997224,
+      "step": 243
+    },
+    {
+      "epoch": 4.440366972477064,
+      "grad_norm": 0.38855600357055664,
+      "learning_rate": 4.396283958372173e-05,
+      "loss": 0.1733,
+      "num_input_tokens_seen": 10036248,
+      "step": 244
+    },
+    {
+      "epoch": 4.458715596330276,
+      "grad_norm": 0.3860638737678528,
+      "learning_rate": 4.391536957168733e-05,
+      "loss": 0.1936,
+      "num_input_tokens_seen": 10070312,
+      "step": 245
+    },
+    {
+      "epoch": 4.477064220183486,
+      "grad_norm": 0.31510865688323975,
+      "learning_rate": 4.386773950556931e-05,
+      "loss": 0.1847,
+      "num_input_tokens_seen": 10114568,
+      "step": 246
+    },
+    {
+      "epoch": 4.495412844036697,
+      "grad_norm": 0.3280925154685974,
+      "learning_rate": 4.381994978839371e-05,
+      "loss": 0.1981,
+      "num_input_tokens_seen": 10150280,
+      "step": 247
+    },
+    {
+      "epoch": 4.513761467889909,
+      "grad_norm": 0.33136090636253357,
+      "learning_rate": 4.377200082453749e-05,
+      "loss": 0.1681,
+      "num_input_tokens_seen": 10194000,
+      "step": 248
+    },
+    {
+      "epoch": 4.532110091743119,
+      "grad_norm": 0.43955501914024353,
+      "learning_rate": 4.372389301972506e-05,
+      "loss": 0.2111,
+      "num_input_tokens_seen": 10232264,
+      "step": 249
+    },
+    {
+      "epoch": 4.5504587155963305,
+      "grad_norm": 0.28938886523246765,
+      "learning_rate": 4.36756267810249e-05,
+      "loss": 0.2307,
+      "num_input_tokens_seen": 10271880,
+      "step": 250
+    },
+    {
+      "epoch": 4.568807339449541,
+      "grad_norm": 0.37232765555381775,
+      "learning_rate": 4.36272025168461e-05,
+      "loss": 0.1609,
+      "num_input_tokens_seen": 10317720,
+      "step": 251
+    },
+    {
+      "epoch": 4.587155963302752,
+      "grad_norm": 0.6248548030853271,
+      "learning_rate": 4.357862063693486e-05,
+      "loss": 0.2456,
+      "num_input_tokens_seen": 10362168,
+      "step": 252
+    },
+    {
+      "epoch": 4.605504587155964,
+      "grad_norm": 0.32828131318092346,
+      "learning_rate": 4.3529881552371096e-05,
+      "loss": 0.3159,
+      "num_input_tokens_seen": 10414312,
+      "step": 253
+    },
+    {
+      "epoch": 4.623853211009174,
+      "grad_norm": 0.4158332049846649,
+      "learning_rate": 4.34809856755649e-05,
+      "loss": 0.2194,
+      "num_input_tokens_seen": 10451800,
+      "step": 254
+    },
+    {
+      "epoch": 4.6422018348623855,
+      "grad_norm": 0.3648194670677185,
+      "learning_rate": 4.34319334202531e-05,
+      "loss": 0.1846,
+      "num_input_tokens_seen": 10496224,
+      "step": 255
+    },
+    {
+      "epoch": 4.660550458715596,
+      "grad_norm": 0.36835575103759766,
+      "learning_rate": 4.3382725201495723e-05,
+      "loss": 0.1906,
+      "num_input_tokens_seen": 10536392,
+      "step": 256
+    },
+    {
+      "epoch": 4.678899082568807,
+      "grad_norm": 0.3501613140106201,
+      "learning_rate": 4.333336143567247e-05,
+      "loss": 0.1793,
+      "num_input_tokens_seen": 10577640,
+      "step": 257
+    },
+    {
+      "epoch": 4.697247706422019,
+      "grad_norm": 0.3431616425514221,
+      "learning_rate": 4.3283842540479264e-05,
+      "loss": 0.1576,
+      "num_input_tokens_seen": 10613376,
+      "step": 258
+    },
+    {
+      "epoch": 4.715596330275229,
+      "grad_norm": 0.3290237784385681,
+      "learning_rate": 4.3234168934924636e-05,
+      "loss": 0.1447,
+      "num_input_tokens_seen": 10647232,
+      "step": 259
+    },
+    {
+      "epoch": 4.73394495412844,
+      "grad_norm": 0.40264153480529785,
+      "learning_rate": 4.318434103932622e-05,
+      "loss": 0.1488,
+      "num_input_tokens_seen": 10696024,
+      "step": 260
+    },
+    {
+      "epoch": 4.752293577981652,
+      "grad_norm": 0.3453703820705414,
+      "learning_rate": 4.313435927530719e-05,
+      "loss": 0.1597,
+      "num_input_tokens_seen": 10730408,
+      "step": 261
+    },
+    {
+      "epoch": 4.770642201834862,
+      "grad_norm": 0.3993653655052185,
+      "learning_rate": 4.30842240657927e-05,
+      "loss": 0.2266,
+      "num_input_tokens_seen": 10764776,
+      "step": 262
+    },
+    {
+      "epoch": 4.7889908256880735,
+      "grad_norm": 0.4080798923969269,
+      "learning_rate": 4.303393583500628e-05,
+      "loss": 0.1562,
+      "num_input_tokens_seen": 10792272,
+      "step": 263
+    },
+    {
+      "epoch": 4.807339449541285,
+      "grad_norm": 0.3028413951396942,
+      "learning_rate": 4.2983495008466276e-05,
+      "loss": 0.1504,
+      "num_input_tokens_seen": 10825240,
+      "step": 264
+    },
+    {
+      "epoch": 4.825688073394495,
+      "grad_norm": 0.3980019688606262,
+      "learning_rate": 4.293290201298223e-05,
+      "loss": 0.1648,
+      "num_input_tokens_seen": 10883592,
+      "step": 265
+    },
+    {
+      "epoch": 4.844036697247707,
+      "grad_norm": 0.42035797238349915,
+      "learning_rate": 4.288215727665129e-05,
+      "loss": 0.1652,
+      "num_input_tokens_seen": 10922640,
+      "step": 266
+    },
+    {
+      "epoch": 4.862385321100917,
+      "grad_norm": 0.37766003608703613,
+      "learning_rate": 4.2831261228854544e-05,
+      "loss": 0.1817,
+      "num_input_tokens_seen": 10967288,
+      "step": 267
+    },
+    {
+      "epoch": 4.8807339449541285,
+      "grad_norm": 0.3495088815689087,
+      "learning_rate": 4.278021430025343e-05,
+      "loss": 0.2066,
+      "num_input_tokens_seen": 11011152,
+      "step": 268
+    },
+    {
+      "epoch": 4.89908256880734,
+      "grad_norm": 0.34071093797683716,
+      "learning_rate": 4.272901692278609e-05,
+      "loss": 0.1522,
+      "num_input_tokens_seen": 11055608,
+      "step": 269
+    },
+    {
+      "epoch": 4.91743119266055,
+      "grad_norm": 0.33437398076057434,
+      "learning_rate": 4.267766952966369e-05,
+      "loss": 0.2201,
+      "num_input_tokens_seen": 11101992,
+      "step": 270
+    },
+    {
+      "epoch": 4.935779816513762,
+      "grad_norm": 0.3340584337711334,
+      "learning_rate": 4.262617255536676e-05,
+      "loss": 0.2777,
+      "num_input_tokens_seen": 11141408,
+      "step": 271
+    },
+    {
+      "epoch": 4.954128440366972,
+      "grad_norm": 0.348679780960083,
+      "learning_rate": 4.257452643564155e-05,
+      "loss": 0.1857,
+      "num_input_tokens_seen": 11185344,
+      "step": 272
+    },
+    {
+      "epoch": 4.972477064220183,
+      "grad_norm": 0.3691309094429016,
+      "learning_rate": 4.2522731607496275e-05,
+      "loss": 0.1653,
+      "num_input_tokens_seen": 11216160,
+      "step": 273
+    },
+    {
+      "epoch": 4.990825688073395,
+      "grad_norm": 0.4692390561103821,
+      "learning_rate": 4.24707885091975e-05,
+      "loss": 0.1761,
+      "num_input_tokens_seen": 11276200,
+      "step": 274
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.4963766932487488,
+      "learning_rate": 4.241869758026638e-05,
+      "loss": 0.1582,
+      "num_input_tokens_seen": 11295440,
+      "step": 275
+    },
+    {
+      "epoch": 5.018348623853211,
+      "grad_norm": 0.37284544110298157,
+      "learning_rate": 4.2366459261474933e-05,
+      "loss": 0.1538,
+      "num_input_tokens_seen": 11336400,
+      "step": 276
+    },
+    {
+      "epoch": 5.036697247706422,
+      "grad_norm": 0.2949577867984772,
+      "learning_rate": 4.231407399484236e-05,
+      "loss": 0.1319,
+      "num_input_tokens_seen": 11374056,
+      "step": 277
+    },
+    {
+      "epoch": 5.055045871559633,
+      "grad_norm": 0.32850027084350586,
+      "learning_rate": 4.226154222363124e-05,
+      "loss": 0.174,
+      "num_input_tokens_seen": 11414776,
+      "step": 278
+    },
+    {
+      "epoch": 5.073394495412844,
+      "grad_norm": 0.3812579810619354,
+      "learning_rate": 4.220886439234385e-05,
+      "loss": 0.1831,
+      "num_input_tokens_seen": 11465456,
+      "step": 279
+    },
+    {
+      "epoch": 5.091743119266055,
+      "grad_norm": 0.3396337032318115,
+      "learning_rate": 4.215604094671835e-05,
+      "loss": 0.1515,
+      "num_input_tokens_seen": 11500184,
+      "step": 280
+    },
+    {
+      "epoch": 5.110091743119266,
+      "grad_norm": 0.35079169273376465,
+      "learning_rate": 4.2103072333725e-05,
+      "loss": 0.1295,
+      "num_input_tokens_seen": 11537112,
+      "step": 281
+    },
+    {
+      "epoch": 5.128440366972477,
+      "grad_norm": 0.3811327815055847,
+      "learning_rate": 4.2049959001562464e-05,
+      "loss": 0.1339,
+      "num_input_tokens_seen": 11579440,
+      "step": 282
+    },
+    {
+      "epoch": 5.146788990825688,
+      "grad_norm": 0.3935602009296417,
+      "learning_rate": 4.199670139965393e-05,
+      "loss": 0.1909,
+      "num_input_tokens_seen": 11643272,
+      "step": 283
+    },
+    {
+      "epoch": 5.165137614678899,
+      "grad_norm": 0.3952607810497284,
+      "learning_rate": 4.194329997864331e-05,
+      "loss": 0.2334,
+      "num_input_tokens_seen": 11677528,
+      "step": 284
+    },
+    {
+      "epoch": 5.18348623853211,
+      "grad_norm": 0.3687472939491272,
+      "learning_rate": 4.188975519039151e-05,
+      "loss": 0.1406,
+      "num_input_tokens_seen": 11727944,
+      "step": 285
+    },
+    {
+      "epoch": 5.201834862385321,
+      "grad_norm": 0.3518407344818115,
+      "learning_rate": 4.183606748797251e-05,
+      "loss": 0.138,
+      "num_input_tokens_seen": 11779568,
+      "step": 286
+    },
+    {
+      "epoch": 5.220183486238533,
+      "grad_norm": 0.4072832763195038,
+      "learning_rate": 4.1782237325669595e-05,
+      "loss": 0.159,
+      "num_input_tokens_seen": 11824600,
+      "step": 287
+    },
+    {
+      "epoch": 5.238532110091743,
+      "grad_norm": 0.386280357837677,
+      "learning_rate": 4.172826515897146e-05,
+      "loss": 0.2517,
+      "num_input_tokens_seen": 11873736,
+      "step": 288
+    },
+    {
+      "epoch": 5.256880733944954,
+      "grad_norm": 0.3576860725879669,
+      "learning_rate": 4.167415144456841e-05,
+      "loss": 0.1349,
+      "num_input_tokens_seen": 11909608,
+      "step": 289
+    },
+    {
+      "epoch": 5.275229357798165,
+      "grad_norm": 0.3952435851097107,
+      "learning_rate": 4.1619896640348445e-05,
+      "loss": 0.1348,
+      "num_input_tokens_seen": 11945440,
+      "step": 290
+    },
+    {
+      "epoch": 5.293577981651376,
+      "grad_norm": 0.3565181493759155,
+      "learning_rate": 4.1565501205393445e-05,
+      "loss": 0.1331,
+      "num_input_tokens_seen": 11985568,
+      "step": 291
+    },
+    {
+      "epoch": 5.3119266055045875,
+      "grad_norm": 0.40558820962905884,
+      "learning_rate": 4.1510965599975196e-05,
+      "loss": 0.2337,
+      "num_input_tokens_seen": 12034320,
+      "step": 292
+    },
+    {
+      "epoch": 5.330275229357798,
+      "grad_norm": 0.36426106095314026,
+      "learning_rate": 4.1456290285551596e-05,
+      "loss": 0.1299,
+      "num_input_tokens_seen": 12070184,
+      "step": 293
+    },
+    {
+      "epoch": 5.348623853211009,
+      "grad_norm": 0.33881404995918274,
+      "learning_rate": 4.140147572476268e-05,
+      "loss": 0.1239,
+      "num_input_tokens_seen": 12111512,
+      "step": 294
+    },
+    {
+      "epoch": 5.36697247706422,
+      "grad_norm": 0.38019630312919617,
+      "learning_rate": 4.1346522381426744e-05,
+      "loss": 0.133,
+      "num_input_tokens_seen": 12156792,
+      "step": 295
+    },
+    {
+      "epoch": 5.385321100917431,
+      "grad_norm": 0.38179296255111694,
+      "learning_rate": 4.129143072053638e-05,
+      "loss": 0.1301,
+      "num_input_tokens_seen": 12185168,
+      "step": 296
+    },
+    {
+      "epoch": 5.4036697247706424,
+      "grad_norm": 0.35054150223731995,
+      "learning_rate": 4.123620120825459e-05,
+      "loss": 0.1298,
+      "num_input_tokens_seen": 12222256,
+      "step": 297
+    },
+    {
+      "epoch": 5.422018348623853,
+      "grad_norm": 0.3463946580886841,
+      "learning_rate": 4.118083431191081e-05,
+      "loss": 0.2088,
+      "num_input_tokens_seen": 12257536,
+      "step": 298
+    },
+    {
+      "epoch": 5.440366972477064,
+      "grad_norm": 0.42584434151649475,
+      "learning_rate": 4.112533049999696e-05,
+      "loss": 0.1062,
+      "num_input_tokens_seen": 12290576,
+      "step": 299
+    },
+    {
+      "epoch": 5.458715596330276,
+      "grad_norm": 0.4341515600681305,
+      "learning_rate": 4.1069690242163484e-05,
+      "loss": 0.1989,
+      "num_input_tokens_seen": 12323416,
+      "step": 300
+    },
+    {
+      "epoch": 5.477064220183486,
+      "grad_norm": 0.4299938380718231,
+      "learning_rate": 4.101391400921538e-05,
+      "loss": 0.1243,
+      "num_input_tokens_seen": 12370264,
+      "step": 301
+    },
+    {
+      "epoch": 5.495412844036697,
+      "grad_norm": 0.4070415198802948,
+      "learning_rate": 4.095800227310821e-05,
+      "loss": 0.2281,
+      "num_input_tokens_seen": 12410568,
+      "step": 302
+    },
+    {
+      "epoch": 5.513761467889909,
+      "grad_norm": 0.4446506202220917,
+      "learning_rate": 4.09019555069441e-05,
+      "loss": 0.1462,
+      "num_input_tokens_seen": 12442880,
+      "step": 303
+    },
+    {
+      "epoch": 5.532110091743119,
+      "grad_norm": 0.36110538244247437,
+      "learning_rate": 4.0845774184967754e-05,
+      "loss": 0.1497,
+      "num_input_tokens_seen": 12487016,
+      "step": 304
+    },
+    {
+      "epoch": 5.5504587155963305,
+      "grad_norm": 0.42756903171539307,
+      "learning_rate": 4.078945878256244e-05,
+      "loss": 0.2082,
+      "num_input_tokens_seen": 12525072,
+      "step": 305
+    },
+    {
+      "epoch": 5.568807339449541,
+      "grad_norm": 0.360649049282074,
+      "learning_rate": 4.073300977624594e-05,
+      "loss": 0.1214,
+      "num_input_tokens_seen": 12555792,
+      "step": 306
+    },
+    {
+      "epoch": 5.587155963302752,
+      "grad_norm": 0.3768391013145447,
+      "learning_rate": 4.067642764366654e-05,
+      "loss": 0.1278,
+      "num_input_tokens_seen": 12601616,
+      "step": 307
+    },
+    {
+      "epoch": 5.605504587155964,
+      "grad_norm": 0.3882285952568054,
+      "learning_rate": 4.0619712863599e-05,
+      "loss": 0.1485,
+      "num_input_tokens_seen": 12634360,
+      "step": 308
+    },
+    {
+      "epoch": 5.623853211009174,
+      "grad_norm": 0.38816162943840027,
+      "learning_rate": 4.0562865915940496e-05,
+      "loss": 0.1221,
+      "num_input_tokens_seen": 12674808,
+      "step": 309
+    },
+    {
+      "epoch": 5.6422018348623855,
+      "grad_norm": 0.41803887486457825,
+      "learning_rate": 4.05058872817065e-05,
+      "loss": 0.1388,
+      "num_input_tokens_seen": 12710864,
+      "step": 310
+    },
+    {
+      "epoch": 5.660550458715596,
+      "grad_norm": 0.3802741467952728,
+      "learning_rate": 4.044877744302683e-05,
+      "loss": 0.1349,
+      "num_input_tokens_seen": 12750920,
+      "step": 311
+    },
+    {
+      "epoch": 5.678899082568807,
+      "grad_norm": 0.48202404379844666,
+      "learning_rate": 4.039153688314145e-05,
+      "loss": 0.1555,
+      "num_input_tokens_seen": 12789488,
+      "step": 312
+    },
+    {
+      "epoch": 5.697247706422019,
+      "grad_norm": 0.3168826103210449,
+      "learning_rate": 4.0334166086396484e-05,
+      "loss": 0.1063,
+      "num_input_tokens_seen": 12831408,
+      "step": 313
+    },
+    {
+      "epoch": 5.715596330275229,
+      "grad_norm": 0.43414828181266785,
+      "learning_rate": 4.0276665538239996e-05,
+      "loss": 0.127,
+      "num_input_tokens_seen": 12872584,
+      "step": 314
+    },
+    {
+      "epoch": 5.73394495412844,
+      "grad_norm": 0.47761547565460205,
+      "learning_rate": 4.021903572521802e-05,
+      "loss": 0.1428,
+      "num_input_tokens_seen": 12910528,
+      "step": 315
+    },
+    {
+      "epoch": 5.752293577981652,
+      "grad_norm": 0.3542017936706543,
+      "learning_rate": 4.0161277134970345e-05,
+      "loss": 0.1279,
+      "num_input_tokens_seen": 12942800,
+      "step": 316
+    },
+    {
+      "epoch": 5.770642201834862,
+      "grad_norm": 0.31866851449012756,
+      "learning_rate": 4.010339025622641e-05,
+      "loss": 0.1459,
+      "num_input_tokens_seen": 12989136,
+      "step": 317
+    },
+    {
+      "epoch": 5.7889908256880735,
+      "grad_norm": 0.44918256998062134,
+      "learning_rate": 4.0045375578801214e-05,
+      "loss": 0.1429,
+      "num_input_tokens_seen": 13035472,
+      "step": 318
+    },
+    {
+      "epoch": 5.807339449541285,
+      "grad_norm": 0.32426726818084717,
+      "learning_rate": 3.99872335935911e-05,
+      "loss": 0.1257,
+      "num_input_tokens_seen": 13074952,
+      "step": 319
+    },
+    {
+      "epoch": 5.825688073394495,
+      "grad_norm": 0.6903991103172302,
+      "learning_rate": 3.9928964792569655e-05,
+      "loss": 0.1807,
+      "num_input_tokens_seen": 13124624,
+      "step": 320
+    },
+    {
+      "epoch": 5.844036697247707,
+      "grad_norm": 0.3665274679660797,
+      "learning_rate": 3.9870569668783536e-05,
+      "loss": 0.1853,
+      "num_input_tokens_seen": 13171464,
+      "step": 321
+    },
+    {
+      "epoch": 5.862385321100917,
+      "grad_norm": 0.41457998752593994,
+      "learning_rate": 3.981204871634827e-05,
+      "loss": 0.214,
+      "num_input_tokens_seen": 13225240,
+      "step": 322
+    },
+    {
+      "epoch": 5.8807339449541285,
+      "grad_norm": 0.4047159254550934,
+      "learning_rate": 3.9753402430444116e-05,
+      "loss": 0.1907,
+      "num_input_tokens_seen": 13275848,
+      "step": 323
+    },
+    {
+      "epoch": 5.89908256880734,
+      "grad_norm": 0.4578211307525635,
+      "learning_rate": 3.969463130731183e-05,
+      "loss": 0.27,
+      "num_input_tokens_seen": 13311096,
+      "step": 324
+    },
+    {
+      "epoch": 5.91743119266055,
+      "grad_norm": 0.42030438780784607,
+      "learning_rate": 3.963573584424852e-05,
+      "loss": 0.1775,
+      "num_input_tokens_seen": 13375064,
+      "step": 325
+    },
+    {
+      "epoch": 5.935779816513762,
+      "grad_norm": 0.3531726598739624,
+      "learning_rate": 3.957671653960337e-05,
+      "loss": 0.1206,
+      "num_input_tokens_seen": 13427072,
+      "step": 326
+    },
+    {
+      "epoch": 5.954128440366972,
+      "grad_norm": 0.41528916358947754,
+      "learning_rate": 3.9517573892773494e-05,
+      "loss": 0.1232,
+      "num_input_tokens_seen": 13454088,
+      "step": 327
+    },
+    {
+      "epoch": 5.972477064220183,
+      "grad_norm": 0.5357607007026672,
+      "learning_rate": 3.945830840419966e-05,
+      "loss": 0.1696,
+      "num_input_tokens_seen": 13486504,
+      "step": 328
+    },
+    {
+      "epoch": 5.990825688073395,
+      "grad_norm": 0.32135313749313354,
+      "learning_rate": 3.9398920575362086e-05,
+      "loss": 0.1055,
+      "num_input_tokens_seen": 13523736,
+      "step": 329
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.5267924666404724,
+      "learning_rate": 3.933941090877615e-05,
+      "loss": 0.2192,
+      "num_input_tokens_seen": 13554528,
+      "step": 330
+    },
+    {
+      "epoch": 6.018348623853211,
+      "grad_norm": 0.38375183939933777,
+      "learning_rate": 3.9279779907988215e-05,
+      "loss": 0.1318,
+      "num_input_tokens_seen": 13587424,
+      "step": 331
+    },
+    {
+      "epoch": 6.036697247706422,
+      "grad_norm": 0.38274672627449036,
+      "learning_rate": 3.9220028077571295e-05,
+      "loss": 0.1136,
+      "num_input_tokens_seen": 13624376,
+      "step": 332
+    },
+    {
+      "epoch": 6.055045871559633,
+      "grad_norm": 0.3571849465370178,
+      "learning_rate": 3.916015592312082e-05,
+      "loss": 0.0891,
+      "num_input_tokens_seen": 13652848,
+      "step": 333
+    },
+    {
+      "epoch": 6.073394495412844,
+      "grad_norm": 0.3958500921726227,
+      "learning_rate": 3.910016395125037e-05,
+      "loss": 0.1477,
+      "num_input_tokens_seen": 13719880,
+      "step": 334
+    },
+    {
+      "epoch": 6.091743119266055,
+      "grad_norm": 0.3360799551010132,
+      "learning_rate": 3.9040052669587325e-05,
+      "loss": 0.0943,
+      "num_input_tokens_seen": 13764864,
+      "step": 335
+    },
+    {
+      "epoch": 6.110091743119266,
+      "grad_norm": 0.4332111179828644,
+      "learning_rate": 3.897982258676867e-05,
+      "loss": 0.1039,
+      "num_input_tokens_seen": 13795048,
+      "step": 336
+    },
+    {
+      "epoch": 6.128440366972477,
+      "grad_norm": 0.4539923071861267,
+      "learning_rate": 3.891947421243661e-05,
+      "loss": 0.1294,
+      "num_input_tokens_seen": 13830064,
+      "step": 337
+    },
+    {
+      "epoch": 6.146788990825688,
+      "grad_norm": 0.43116849660873413,
+      "learning_rate": 3.885900805723429e-05,
+      "loss": 0.1233,
+      "num_input_tokens_seen": 13870392,
+      "step": 338
+    },
+    {
+      "epoch": 6.165137614678899,
+      "grad_norm": 0.40204253792762756,
+      "learning_rate": 3.879842463280145e-05,
+      "loss": 0.1252,
+      "num_input_tokens_seen": 13913144,
+      "step": 339
+    },
+    {
+      "epoch": 6.18348623853211,
+      "grad_norm": 0.4300351142883301,
+      "learning_rate": 3.873772445177015e-05,
+      "loss": 0.1191,
+      "num_input_tokens_seen": 13950384,
+      "step": 340
+    },
+    {
+      "epoch": 6.201834862385321,
+      "grad_norm": 0.3162730038166046,
+      "learning_rate": 3.8676908027760364e-05,
+      "loss": 0.1028,
+      "num_input_tokens_seen": 13990464,
+      "step": 341
+    },
+    {
+      "epoch": 6.220183486238533,
+      "grad_norm": 0.4316507875919342,
+      "learning_rate": 3.861597587537568e-05,
+      "loss": 0.0996,
+      "num_input_tokens_seen": 14029080,
+      "step": 342
+    },
+    {
+      "epoch": 6.238532110091743,
+      "grad_norm": 0.34346097707748413,
+      "learning_rate": 3.855492851019893e-05,
+      "loss": 0.172,
+      "num_input_tokens_seen": 14069824,
+      "step": 343
+    },
+    {
+      "epoch": 6.256880733944954,
+      "grad_norm": 0.41037800908088684,
+      "learning_rate": 3.8493766448787825e-05,
+      "loss": 0.1244,
+      "num_input_tokens_seen": 14107616,
+      "step": 344
+    },
+    {
+      "epoch": 6.275229357798165,
+      "grad_norm": 0.5019925236701965,
+      "learning_rate": 3.84324902086706e-05,
+      "loss": 0.1584,
+      "num_input_tokens_seen": 14146224,
+      "step": 345
+    },
+    {
+      "epoch": 6.293577981651376,
+      "grad_norm": 0.40119829773902893,
+      "learning_rate": 3.837110030834161e-05,
+      "loss": 0.1154,
+      "num_input_tokens_seen": 14183880,
+      "step": 346
+    },
+    {
+      "epoch": 6.3119266055045875,
+      "grad_norm": 0.37845325469970703,
+      "learning_rate": 3.830959726725697e-05,
+      "loss": 0.098,
+      "num_input_tokens_seen": 14214632,
+      "step": 347
+    },
+    {
+      "epoch": 6.330275229357798,
+      "grad_norm": 0.338489294052124,
+      "learning_rate": 3.824798160583012e-05,
+      "loss": 0.2212,
+      "num_input_tokens_seen": 14263168,
+      "step": 348
+    },
+    {
+      "epoch": 6.348623853211009,
+      "grad_norm": 0.4595409631729126,
+      "learning_rate": 3.81862538454275e-05,
+      "loss": 0.1423,
+      "num_input_tokens_seen": 14303680,
+      "step": 349
+    },
+    {
+      "epoch": 6.36697247706422,
+      "grad_norm": 0.38874781131744385,
+      "learning_rate": 3.8124414508364e-05,
+      "loss": 0.1119,
+      "num_input_tokens_seen": 14347672,
+      "step": 350
+    },
+    {
+      "epoch": 6.385321100917431,
+      "grad_norm": 0.3326081931591034,
+      "learning_rate": 3.8062464117898724e-05,
+      "loss": 0.1068,
+      "num_input_tokens_seen": 14394848,
+      "step": 351
+    },
+    {
+      "epoch": 6.4036697247706424,
+      "grad_norm": 0.40956059098243713,
+      "learning_rate": 3.8000403198230387e-05,
+      "loss": 0.1051,
+      "num_input_tokens_seen": 14438856,
+      "step": 352
+    },
+    {
+      "epoch": 6.422018348623853,
+      "grad_norm": 0.353677362203598,
+      "learning_rate": 3.7938232274493e-05,
+      "loss": 0.2208,
+      "num_input_tokens_seen": 14498032,
+      "step": 353
+    },
+    {
+      "epoch": 6.440366972477064,
+      "grad_norm": 0.45626017451286316,
+      "learning_rate": 3.787595187275136e-05,
+      "loss": 0.1131,
+      "num_input_tokens_seen": 14534256,
+      "step": 354
+    },
+    {
+      "epoch": 6.458715596330276,
+      "grad_norm": 0.4512569308280945,
+      "learning_rate": 3.781356251999663e-05,
+      "loss": 0.1207,
+      "num_input_tokens_seen": 14575016,
+      "step": 355
+    },
+    {
+      "epoch": 6.477064220183486,
+      "grad_norm": 0.3183128237724304,
+      "learning_rate": 3.775106474414188e-05,
+      "loss": 0.0887,
+      "num_input_tokens_seen": 14627272,
+      "step": 356
+    },
+    {
+      "epoch": 6.495412844036697,
+      "grad_norm": 0.36176785826683044,
+      "learning_rate": 3.7688459074017606e-05,
+      "loss": 0.1022,
+      "num_input_tokens_seen": 14658696,
+      "step": 357
+    },
+    {
+      "epoch": 6.513761467889909,
+      "grad_norm": 0.362131804227829,
+      "learning_rate": 3.762574603936725e-05,
+      "loss": 0.1582,
+      "num_input_tokens_seen": 14702208,
+      "step": 358
+    },
+    {
+      "epoch": 6.532110091743119,
+      "grad_norm": 0.27560174465179443,
+      "learning_rate": 3.756292617084275e-05,
+      "loss": 0.0855,
+      "num_input_tokens_seen": 14757376,
+      "step": 359
+    },
+    {
+      "epoch": 6.5504587155963305,
+      "grad_norm": 0.4014551043510437,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.1923,
+      "num_input_tokens_seen": 14796592,
+      "step": 360
+    },
+    {
+      "epoch": 6.568807339449541,
+      "grad_norm": 0.38982701301574707,
+      "learning_rate": 3.7436968059294414e-05,
+      "loss": 0.1125,
+      "num_input_tokens_seen": 14834656,
+      "step": 361
+    },
+    {
+      "epoch": 6.587155963302752,
+      "grad_norm": 0.3973703384399414,
+      "learning_rate": 3.7373830882076354e-05,
+      "loss": 0.1947,
+      "num_input_tokens_seen": 14869528,
+      "step": 362
+    },
+    {
+      "epoch": 6.605504587155964,
+      "grad_norm": 0.43253588676452637,
+      "learning_rate": 3.731058900258668e-05,
+      "loss": 0.1173,
+      "num_input_tokens_seen": 14909720,
+      "step": 363
+    },
+    {
+      "epoch": 6.623853211009174,
+      "grad_norm": 0.4936161935329437,
+      "learning_rate": 3.7247242955952175e-05,
+      "loss": 0.1523,
+      "num_input_tokens_seen": 14952704,
+      "step": 364
+    },
+    {
+      "epoch": 6.6422018348623855,
+      "grad_norm": 0.43221941590309143,
+      "learning_rate": 3.718379327818106e-05,
+      "loss": 0.1313,
+      "num_input_tokens_seen": 14989640,
+      "step": 365
+    },
+    {
+      "epoch": 6.660550458715596,
+      "grad_norm": 0.3976401686668396,
+      "learning_rate": 3.712024050615843e-05,
+      "loss": 0.0985,
+      "num_input_tokens_seen": 15028200,
+      "step": 366
+    },
+    {
+      "epoch": 6.678899082568807,
+      "grad_norm": 0.3843483030796051,
+      "learning_rate": 3.705658517764172e-05,
+      "loss": 0.1905,
+      "num_input_tokens_seen": 15075960,
+      "step": 367
+    },
+    {
+      "epoch": 6.697247706422019,
+      "grad_norm": 0.45637550950050354,
+      "learning_rate": 3.699282783125616e-05,
+      "loss": 0.1166,
+      "num_input_tokens_seen": 15113096,
+      "step": 368
+    },
+    {
+      "epoch": 6.715596330275229,
+      "grad_norm": 0.3767227232456207,
+      "learning_rate": 3.692896900649021e-05,
+      "loss": 0.138,
+      "num_input_tokens_seen": 15162432,
+      "step": 369
+    },
+    {
+      "epoch": 6.73394495412844,
+      "grad_norm": 0.3972814977169037,
+      "learning_rate": 3.686500924369101e-05,
+      "loss": 0.1612,
+      "num_input_tokens_seen": 15203656,
+      "step": 370
+    },
+    {
+      "epoch": 6.752293577981652,
+      "grad_norm": 0.39401331543922424,
+      "learning_rate": 3.680094908405978e-05,
+      "loss": 0.1131,
+      "num_input_tokens_seen": 15251056,
+      "step": 371
+    },
+    {
+      "epoch": 6.770642201834862,
+      "grad_norm": 0.5119125247001648,
+      "learning_rate": 3.673678906964727e-05,
+      "loss": 0.1264,
+      "num_input_tokens_seen": 15283552,
+      "step": 372
+    },
+    {
+      "epoch": 6.7889908256880735,
+      "grad_norm": 0.38780996203422546,
+      "learning_rate": 3.6672529743349146e-05,
+      "loss": 0.1167,
+      "num_input_tokens_seen": 15338944,
+      "step": 373
+    },
+    {
+      "epoch": 6.807339449541285,
+      "grad_norm": 0.45605945587158203,
+      "learning_rate": 3.660817164890143e-05,
+      "loss": 0.131,
+      "num_input_tokens_seen": 15369936,
+      "step": 374
+    },
+    {
+      "epoch": 6.825688073394495,
+      "grad_norm": 0.4043484926223755,
+      "learning_rate": 3.654371533087586e-05,
+      "loss": 0.095,
+      "num_input_tokens_seen": 15411576,
+      "step": 375
+    },
+    {
+      "epoch": 6.844036697247707,
+      "grad_norm": 0.38716065883636475,
+      "learning_rate": 3.6479161334675296e-05,
+      "loss": 0.0937,
+      "num_input_tokens_seen": 15457520,
+      "step": 376
+    },
+    {
+      "epoch": 6.862385321100917,
+      "grad_norm": 0.38785362243652344,
+      "learning_rate": 3.641451020652914e-05,
+      "loss": 0.1017,
+      "num_input_tokens_seen": 15485536,
+      "step": 377
+    },
+    {
+      "epoch": 6.8807339449541285,
+      "grad_norm": 0.3736408054828644,
+      "learning_rate": 3.634976249348867e-05,
+      "loss": 0.1075,
+      "num_input_tokens_seen": 15526704,
+      "step": 378
+    },
+    {
+      "epoch": 6.89908256880734,
+      "grad_norm": 0.5013216733932495,
+      "learning_rate": 3.6284918743422425e-05,
+      "loss": 0.132,
+      "num_input_tokens_seen": 15567344,
+      "step": 379
+    },
+    {
+      "epoch": 6.91743119266055,
+      "grad_norm": 0.5258253216743469,
+      "learning_rate": 3.621997950501156e-05,
+      "loss": 0.1435,
+      "num_input_tokens_seen": 15608408,
+      "step": 380
+    },
+    {
+      "epoch": 6.935779816513762,
+      "grad_norm": 0.3486093282699585,
+      "learning_rate": 3.615494532774522e-05,
+      "loss": 0.0958,
+      "num_input_tokens_seen": 15667496,
+      "step": 381
+    },
+    {
+      "epoch": 6.954128440366972,
+      "grad_norm": 0.39199239015579224,
+      "learning_rate": 3.6089816761915906e-05,
+      "loss": 0.1465,
+      "num_input_tokens_seen": 15715496,
+      "step": 382
+    },
+    {
+      "epoch": 6.972477064220183,
+      "grad_norm": 0.3722994923591614,
+      "learning_rate": 3.602459435861475e-05,
+      "loss": 0.1111,
+      "num_input_tokens_seen": 15755688,
+      "step": 383
+    },
+    {
+      "epoch": 6.990825688073395,
+      "grad_norm": 0.41041186451911926,
+      "learning_rate": 3.5959278669726935e-05,
+      "loss": 0.1084,
+      "num_input_tokens_seen": 15798304,
+      "step": 384
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.6807722449302673,
+      "learning_rate": 3.589387024792699e-05,
+      "loss": 0.1234,
+      "num_input_tokens_seen": 15813616,
+      "step": 385
+    },
+    {
+      "epoch": 7.018348623853211,
+      "grad_norm": 0.3617340922355652,
+      "learning_rate": 3.582836964667408e-05,
+      "loss": 0.1167,
+      "num_input_tokens_seen": 15855256,
+      "step": 386
+    },
+    {
+      "epoch": 7.036697247706422,
+      "grad_norm": 0.36471137404441833,
+      "learning_rate": 3.576277742020738e-05,
+      "loss": 0.0829,
+      "num_input_tokens_seen": 15887512,
+      "step": 387
+    },
+    {
+      "epoch": 7.055045871559633,
+      "grad_norm": 0.8828345537185669,
+      "learning_rate": 3.569709412354136e-05,
+      "loss": 0.1213,
+      "num_input_tokens_seen": 15934400,
+      "step": 388
+    },
+    {
+      "epoch": 7.073394495412844,
+      "grad_norm": 0.4129248261451721,
+      "learning_rate": 3.563132031246108e-05,
+      "loss": 0.0914,
+      "num_input_tokens_seen": 15969904,
+      "step": 389
+    },
+    {
+      "epoch": 7.091743119266055,
+      "grad_norm": 0.38659873604774475,
+      "learning_rate": 3.556545654351749e-05,
+      "loss": 0.09,
+      "num_input_tokens_seen": 16010760,
+      "step": 390
+    },
+    {
+      "epoch": 7.110091743119266,
+      "grad_norm": 0.46532171964645386,
+      "learning_rate": 3.549950337402274e-05,
+      "loss": 0.0832,
+      "num_input_tokens_seen": 16059680,
+      "step": 391
+    },
+    {
+      "epoch": 7.128440366972477,
+      "grad_norm": 0.3775540888309479,
+      "learning_rate": 3.543346136204545e-05,
+      "loss": 0.079,
+      "num_input_tokens_seen": 16095992,
+      "step": 392
+    },
+    {
+      "epoch": 7.146788990825688,
+      "grad_norm": 0.5437320470809937,
+      "learning_rate": 3.536733106640598e-05,
+      "loss": 0.1,
+      "num_input_tokens_seen": 16135568,
+      "step": 393
+    },
+    {
+      "epoch": 7.165137614678899,
+      "grad_norm": 0.3908531963825226,
+      "learning_rate": 3.5301113046671714e-05,
+      "loss": 0.0962,
+      "num_input_tokens_seen": 16165016,
+      "step": 394
+    },
+    {
+      "epoch": 7.18348623853211,
+      "grad_norm": 0.40789636969566345,
+      "learning_rate": 3.523480786315231e-05,
+      "loss": 0.0923,
+      "num_input_tokens_seen": 16208344,
+      "step": 395
+    },
+    {
+      "epoch": 7.201834862385321,
+      "grad_norm": 0.4682011902332306,
+      "learning_rate": 3.516841607689501e-05,
+      "loss": 0.101,
+      "num_input_tokens_seen": 16244312,
+      "step": 396
+    },
+    {
+      "epoch": 7.220183486238533,
+      "grad_norm": 0.5206179618835449,
+      "learning_rate": 3.5101938249679794e-05,
+      "loss": 0.1124,
+      "num_input_tokens_seen": 16283848,
+      "step": 397
+    },
+    {
+      "epoch": 7.238532110091743,
+      "grad_norm": 0.5803564786911011,
+      "learning_rate": 3.503537494401473e-05,
+      "loss": 0.1054,
+      "num_input_tokens_seen": 16336288,
+      "step": 398
+    },
+    {
+      "epoch": 7.256880733944954,
+      "grad_norm": 0.44892576336860657,
+      "learning_rate": 3.496872672313116e-05,
+      "loss": 0.1781,
+      "num_input_tokens_seen": 16375448,
+      "step": 399
+    },
+    {
+      "epoch": 7.275229357798165,
+      "grad_norm": 0.42554035782814026,
+      "learning_rate": 3.490199415097892e-05,
+      "loss": 0.0935,
+      "num_input_tokens_seen": 16412608,
+      "step": 400
+    },
+    {
+      "epoch": 7.293577981651376,
+      "grad_norm": 0.3525730073451996,
+      "learning_rate": 3.483517779222163e-05,
+      "loss": 0.0944,
+      "num_input_tokens_seen": 16455904,
+      "step": 401
+    },
+    {
+      "epoch": 7.3119266055045875,
+      "grad_norm": 0.3382134735584259,
+      "learning_rate": 3.476827821223184e-05,
+      "loss": 0.0878,
+      "num_input_tokens_seen": 16496824,
+      "step": 402
+    },
+    {
+      "epoch": 7.330275229357798,
+      "grad_norm": 0.38020089268684387,
+      "learning_rate": 3.4701295977086324e-05,
+      "loss": 0.0911,
+      "num_input_tokens_seen": 16529344,
+      "step": 403
+    },
+    {
+      "epoch": 7.348623853211009,
+      "grad_norm": 0.44909772276878357,
+      "learning_rate": 3.463423165356121e-05,
+      "loss": 0.1047,
+      "num_input_tokens_seen": 16563664,
+      "step": 404
+    },
+    {
+      "epoch": 7.36697247706422,
+      "grad_norm": 0.3612752854824066,
+      "learning_rate": 3.456708580912725e-05,
+      "loss": 0.1228,
+      "num_input_tokens_seen": 16607272,
+      "step": 405
+    },
+    {
+      "epoch": 7.385321100917431,
+      "grad_norm": 0.5135468244552612,
+      "learning_rate": 3.449985901194498e-05,
+      "loss": 0.1801,
+      "num_input_tokens_seen": 16646040,
+      "step": 406
+    },
+    {
+      "epoch": 7.4036697247706424,
+      "grad_norm": 0.33949241042137146,
+      "learning_rate": 3.443255183085993e-05,
+      "loss": 0.0883,
+      "num_input_tokens_seen": 16693544,
+      "step": 407
+    },
+    {
+      "epoch": 7.422018348623853,
+      "grad_norm": 0.3560762405395508,
+      "learning_rate": 3.436516483539781e-05,
+      "loss": 0.083,
+      "num_input_tokens_seen": 16742128,
+      "step": 408
+    },
+    {
+      "epoch": 7.440366972477064,
+      "grad_norm": 0.32824087142944336,
+      "learning_rate": 3.4297698595759664e-05,
+      "loss": 0.1247,
+      "num_input_tokens_seen": 16792000,
+      "step": 409
+    },
+    {
+      "epoch": 7.458715596330276,
+      "grad_norm": 0.44831863045692444,
+      "learning_rate": 3.423015368281711e-05,
+      "loss": 0.0948,
+      "num_input_tokens_seen": 16824312,
+      "step": 410
+    },
+    {
+      "epoch": 7.477064220183486,
+      "grad_norm": 0.4942893981933594,
+      "learning_rate": 3.4162530668107434e-05,
+      "loss": 0.0993,
+      "num_input_tokens_seen": 16856344,
+      "step": 411
+    },
+    {
+      "epoch": 7.495412844036697,
+      "grad_norm": 0.3783268332481384,
+      "learning_rate": 3.409483012382879e-05,
+      "loss": 0.1273,
+      "num_input_tokens_seen": 16898256,
+      "step": 412
+    },
+    {
+      "epoch": 7.513761467889909,
+      "grad_norm": 0.5532792210578918,
+      "learning_rate": 3.402705262283537e-05,
+      "loss": 0.1171,
+      "num_input_tokens_seen": 16939304,
+      "step": 413
+    },
+    {
+      "epoch": 7.532110091743119,
+      "grad_norm": 0.3667624592781067,
+      "learning_rate": 3.39591987386325e-05,
+      "loss": 0.1753,
+      "num_input_tokens_seen": 16978632,
+      "step": 414
+    },
+    {
+      "epoch": 7.5504587155963305,
+      "grad_norm": 0.42965346574783325,
+      "learning_rate": 3.389126904537192e-05,
+      "loss": 0.1567,
+      "num_input_tokens_seen": 17016552,
+      "step": 415
+    },
+    {
+      "epoch": 7.568807339449541,
+      "grad_norm": 0.41699931025505066,
+      "learning_rate": 3.382326411784672e-05,
+      "loss": 0.0941,
+      "num_input_tokens_seen": 17067688,
+      "step": 416
+    },
+    {
+      "epoch": 7.587155963302752,
+      "grad_norm": 0.3647185266017914,
+      "learning_rate": 3.375518453148669e-05,
+      "loss": 0.0961,
+      "num_input_tokens_seen": 17126008,
+      "step": 417
+    },
+    {
+      "epoch": 7.605504587155964,
+      "grad_norm": 0.5344395041465759,
+      "learning_rate": 3.3687030862353286e-05,
+      "loss": 0.1974,
+      "num_input_tokens_seen": 17165800,
+      "step": 418
+    },
+    {
+      "epoch": 7.623853211009174,
+      "grad_norm": 0.38556504249572754,
+      "learning_rate": 3.361880368713486e-05,
+      "loss": 0.094,
+      "num_input_tokens_seen": 17199832,
+      "step": 419
+    },
+    {
+      "epoch": 7.6422018348623855,
+      "grad_norm": 0.3512709438800812,
+      "learning_rate": 3.355050358314172e-05,
+      "loss": 0.0717,
+      "num_input_tokens_seen": 17237952,
+      "step": 420
+    },
+    {
+      "epoch": 7.660550458715596,
+      "grad_norm": 0.35488539934158325,
+      "learning_rate": 3.348213112830128e-05,
+      "loss": 0.0943,
+      "num_input_tokens_seen": 17278384,
+      "step": 421
+    },
+    {
+      "epoch": 7.678899082568807,
+      "grad_norm": 0.4220011234283447,
+      "learning_rate": 3.3413686901153165e-05,
+      "loss": 0.0814,
+      "num_input_tokens_seen": 17309544,
+      "step": 422
+    },
+    {
+      "epoch": 7.697247706422019,
+      "grad_norm": 0.42773446440696716,
+      "learning_rate": 3.3345171480844275e-05,
+      "loss": 0.1192,
+      "num_input_tokens_seen": 17344896,
+      "step": 423
+    },
+    {
+      "epoch": 7.715596330275229,
+      "grad_norm": 0.30265432596206665,
+      "learning_rate": 3.327658544712395e-05,
+      "loss": 0.104,
+      "num_input_tokens_seen": 17413176,
+      "step": 424
+    },
+    {
+      "epoch": 7.73394495412844,
+      "grad_norm": 0.38354116678237915,
+      "learning_rate": 3.3207929380339034e-05,
+      "loss": 0.0965,
+      "num_input_tokens_seen": 17450416,
+      "step": 425
+    },
+    {
+      "epoch": 7.752293577981652,
+      "grad_norm": 0.33634060621261597,
+      "learning_rate": 3.313920386142892e-05,
+      "loss": 0.0775,
+      "num_input_tokens_seen": 17491256,
+      "step": 426
+    },
+    {
+      "epoch": 7.770642201834862,
+      "grad_norm": 0.4931448698043823,
+      "learning_rate": 3.3070409471920726e-05,
+      "loss": 0.1651,
+      "num_input_tokens_seen": 17521600,
+      "step": 427
+    },
+    {
+      "epoch": 7.7889908256880735,
+      "grad_norm": 0.4020994007587433,
+      "learning_rate": 3.3001546793924285e-05,
+      "loss": 0.1071,
+      "num_input_tokens_seen": 17564288,
+      "step": 428
+    },
+    {
+      "epoch": 7.807339449541285,
+      "grad_norm": 0.47053155303001404,
+      "learning_rate": 3.293261641012731e-05,
+      "loss": 0.0987,
+      "num_input_tokens_seen": 17604304,
+      "step": 429
+    },
+    {
+      "epoch": 7.825688073394495,
+      "grad_norm": 0.6638731956481934,
+      "learning_rate": 3.2863618903790346e-05,
+      "loss": 0.1743,
+      "num_input_tokens_seen": 17641384,
+      "step": 430
+    },
+    {
+      "epoch": 7.844036697247707,
+      "grad_norm": 0.4213615357875824,
+      "learning_rate": 3.279455485874195e-05,
+      "loss": 0.0854,
+      "num_input_tokens_seen": 17680560,
+      "step": 431
+    },
+    {
+      "epoch": 7.862385321100917,
+      "grad_norm": 0.47264209389686584,
+      "learning_rate": 3.272542485937369e-05,
+      "loss": 0.0949,
+      "num_input_tokens_seen": 17715312,
+      "step": 432
+    },
+    {
+      "epoch": 7.8807339449541285,
+      "grad_norm": 0.3375228941440582,
+      "learning_rate": 3.26562294906352e-05,
+      "loss": 0.1377,
+      "num_input_tokens_seen": 17766056,
+      "step": 433
+    },
+    {
+      "epoch": 7.89908256880734,
+      "grad_norm": 0.4427188038825989,
+      "learning_rate": 3.2586969338029274e-05,
+      "loss": 0.1035,
+      "num_input_tokens_seen": 17809040,
+      "step": 434
+    },
+    {
+      "epoch": 7.91743119266055,
+      "grad_norm": 0.3992091119289398,
+      "learning_rate": 3.251764498760683e-05,
+      "loss": 0.0844,
+      "num_input_tokens_seen": 17856896,
+      "step": 435
+    },
+    {
+      "epoch": 7.935779816513762,
+      "grad_norm": 0.4064525365829468,
+      "learning_rate": 3.244825702596205e-05,
+      "loss": 0.1295,
+      "num_input_tokens_seen": 17917200,
+      "step": 436
+    },
+    {
+      "epoch": 7.954128440366972,
+      "grad_norm": 0.33894112706184387,
+      "learning_rate": 3.237880604022735e-05,
+      "loss": 0.0662,
+      "num_input_tokens_seen": 17967840,
+      "step": 437
+    },
+    {
+      "epoch": 7.972477064220183,
+      "grad_norm": 0.3915395140647888,
+      "learning_rate": 3.230929261806842e-05,
+      "loss": 0.087,
+      "num_input_tokens_seen": 18001712,
+      "step": 438
+    },
+    {
+      "epoch": 7.990825688073395,
+      "grad_norm": 0.4242343008518219,
+      "learning_rate": 3.223971734767928e-05,
+      "loss": 0.0808,
+      "num_input_tokens_seen": 18044712,
+      "step": 439
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.5116259455680847,
+      "learning_rate": 3.217008081777726e-05,
+      "loss": 0.09,
+      "num_input_tokens_seen": 18072704,
+      "step": 440
+    },
+    {
+      "epoch": 8.01834862385321,
+      "grad_norm": 0.3611031174659729,
+      "learning_rate": 3.210038361759807e-05,
+      "loss": 0.0775,
+      "num_input_tokens_seen": 18103680,
+      "step": 441
+    },
+    {
+      "epoch": 8.036697247706423,
+      "grad_norm": 0.2993970513343811,
+      "learning_rate": 3.203062633689077e-05,
+      "loss": 0.0667,
+      "num_input_tokens_seen": 18133248,
+      "step": 442
+    },
+    {
+      "epoch": 8.055045871559633,
+      "grad_norm": 0.3495657444000244,
+      "learning_rate": 3.1960809565912794e-05,
+      "loss": 0.0616,
+      "num_input_tokens_seen": 18185672,
+      "step": 443
+    },
+    {
+      "epoch": 8.073394495412844,
+      "grad_norm": 0.41082751750946045,
+      "learning_rate": 3.1890933895424976e-05,
+      "loss": 0.0909,
+      "num_input_tokens_seen": 18228240,
+      "step": 444
+    },
+    {
+      "epoch": 8.091743119266056,
+      "grad_norm": 0.3840062618255615,
+      "learning_rate": 3.182099991668653e-05,
+      "loss": 0.0688,
+      "num_input_tokens_seen": 18271416,
+      "step": 445
+    },
+    {
+      "epoch": 8.110091743119266,
+      "grad_norm": 0.385258287191391,
+      "learning_rate": 3.1751008221450025e-05,
+      "loss": 0.1181,
+      "num_input_tokens_seen": 18314208,
+      "step": 446
+    },
+    {
+      "epoch": 8.128440366972477,
+      "grad_norm": 0.3149656057357788,
+      "learning_rate": 3.168095940195642e-05,
+      "loss": 0.1408,
+      "num_input_tokens_seen": 18372552,
+      "step": 447
+    },
+    {
+      "epoch": 8.146788990825687,
+      "grad_norm": 0.3338538408279419,
+      "learning_rate": 3.161085405093006e-05,
+      "loss": 0.0613,
+      "num_input_tokens_seen": 18418048,
+      "step": 448
+    },
+    {
+      "epoch": 8.1651376146789,
+      "grad_norm": 0.35740089416503906,
+      "learning_rate": 3.154069276157359e-05,
+      "loss": 0.071,
+      "num_input_tokens_seen": 18459656,
+      "step": 449
+    },
+    {
+      "epoch": 8.18348623853211,
+      "grad_norm": 0.44650277495384216,
+      "learning_rate": 3.147047612756302e-05,
+      "loss": 0.0781,
+      "num_input_tokens_seen": 18489056,
+      "step": 450
+    },
+    {
+      "epoch": 8.20183486238532,
+      "grad_norm": 0.38402098417282104,
+      "learning_rate": 3.140020474304265e-05,
+      "loss": 0.1724,
+      "num_input_tokens_seen": 18536296,
+      "step": 451
+    },
+    {
+      "epoch": 8.220183486238533,
+      "grad_norm": 0.3836778700351715,
+      "learning_rate": 3.132987920262005e-05,
+      "loss": 0.072,
+      "num_input_tokens_seen": 18569984,
+      "step": 452
+    },
+    {
+      "epoch": 8.238532110091743,
+      "grad_norm": 0.47142115235328674,
+      "learning_rate": 3.125950010136104e-05,
+      "loss": 0.0831,
+      "num_input_tokens_seen": 18600496,
+      "step": 453
+    },
+    {
+      "epoch": 8.256880733944953,
+      "grad_norm": 0.3867137134075165,
+      "learning_rate": 3.118906803478465e-05,
+      "loss": 0.0763,
+      "num_input_tokens_seen": 18644784,
+      "step": 454
+    },
+    {
+      "epoch": 8.275229357798166,
+      "grad_norm": 0.344103068113327,
+      "learning_rate": 3.11185835988581e-05,
+      "loss": 0.0693,
+      "num_input_tokens_seen": 18682928,
+      "step": 455
+    },
+    {
+      "epoch": 8.293577981651376,
+      "grad_norm": 0.3809478282928467,
+      "learning_rate": 3.104804738999169e-05,
+      "loss": 0.069,
+      "num_input_tokens_seen": 18735736,
+      "step": 456
+    },
+    {
+      "epoch": 8.311926605504587,
+      "grad_norm": 0.433005690574646,
+      "learning_rate": 3.097746000503386e-05,
+      "loss": 0.0716,
+      "num_input_tokens_seen": 18779416,
+      "step": 457
+    },
+    {
+      "epoch": 8.330275229357799,
+      "grad_norm": 0.4366289973258972,
+      "learning_rate": 3.090682204126604e-05,
+      "loss": 0.081,
+      "num_input_tokens_seen": 18813088,
+      "step": 458
+    },
+    {
+      "epoch": 8.34862385321101,
+      "grad_norm": 0.44695258140563965,
+      "learning_rate": 3.083613409639764e-05,
+      "loss": 0.088,
+      "num_input_tokens_seen": 18838800,
+      "step": 459
+    },
+    {
+      "epoch": 8.36697247706422,
+      "grad_norm": 0.3648586571216583,
+      "learning_rate": 3.076539676856101e-05,
+      "loss": 0.0616,
+      "num_input_tokens_seen": 18882128,
+      "step": 460
+    },
+    {
+      "epoch": 8.385321100917432,
+      "grad_norm": 0.35021987557411194,
+      "learning_rate": 3.069461065630634e-05,
+      "loss": 0.0685,
+      "num_input_tokens_seen": 18927120,
+      "step": 461
+    },
+    {
+      "epoch": 8.403669724770642,
+      "grad_norm": 0.4578837752342224,
+      "learning_rate": 3.062377635859663e-05,
+      "loss": 0.1019,
+      "num_input_tokens_seen": 18968240,
+      "step": 462
+    },
+    {
+      "epoch": 8.422018348623853,
+      "grad_norm": 0.39133012294769287,
+      "learning_rate": 3.0552894474802584e-05,
+      "loss": 0.1115,
+      "num_input_tokens_seen": 19013248,
+      "step": 463
+    },
+    {
+      "epoch": 8.440366972477065,
+      "grad_norm": 0.3527860939502716,
+      "learning_rate": 3.048196560469758e-05,
+      "loss": 0.0648,
+      "num_input_tokens_seen": 19058904,
+      "step": 464
+    },
+    {
+      "epoch": 8.458715596330276,
+      "grad_norm": 0.3016138970851898,
+      "learning_rate": 3.0410990348452573e-05,
+      "loss": 0.0678,
+      "num_input_tokens_seen": 19103128,
+      "step": 465
+    },
+    {
+      "epoch": 8.477064220183486,
+      "grad_norm": 0.4393775761127472,
+      "learning_rate": 3.0339969306631005e-05,
+      "loss": 0.1322,
+      "num_input_tokens_seen": 19144648,
+      "step": 466
+    },
+    {
+      "epoch": 8.495412844036696,
+      "grad_norm": 0.4702647626399994,
+      "learning_rate": 3.0268903080183743e-05,
+      "loss": 0.0935,
+      "num_input_tokens_seen": 19182568,
+      "step": 467
+    },
+    {
+      "epoch": 8.513761467889909,
+      "grad_norm": 0.44502392411231995,
+      "learning_rate": 3.0197792270443982e-05,
+      "loss": 0.1373,
+      "num_input_tokens_seen": 19218336,
+      "step": 468
+    },
+    {
+      "epoch": 8.53211009174312,
+      "grad_norm": 0.4069594144821167,
+      "learning_rate": 3.0126637479122193e-05,
+      "loss": 0.1189,
+      "num_input_tokens_seen": 19258632,
+      "step": 469
+    },
+    {
+      "epoch": 8.55045871559633,
+      "grad_norm": 0.3544933795928955,
+      "learning_rate": 3.0055439308300952e-05,
+      "loss": 0.0683,
+      "num_input_tokens_seen": 19296928,
+      "step": 470
+    },
+    {
+      "epoch": 8.568807339449542,
+      "grad_norm": 0.4445359408855438,
+      "learning_rate": 2.9984198360429932e-05,
+      "loss": 0.0766,
+      "num_input_tokens_seen": 19338416,
+      "step": 471
+    },
+    {
+      "epoch": 8.587155963302752,
+      "grad_norm": 0.3744385838508606,
+      "learning_rate": 2.9912915238320754e-05,
+      "loss": 0.1101,
+      "num_input_tokens_seen": 19392040,
+      "step": 472
+    },
+    {
+      "epoch": 8.605504587155963,
+      "grad_norm": 0.4125937521457672,
+      "learning_rate": 2.9841590545141906e-05,
+      "loss": 0.0811,
+      "num_input_tokens_seen": 19430360,
+      "step": 473
+    },
+    {
+      "epoch": 8.623853211009175,
+      "grad_norm": 0.33389994502067566,
+      "learning_rate": 2.9770224884413623e-05,
+      "loss": 0.0665,
+      "num_input_tokens_seen": 19487640,
+      "step": 474
+    },
+    {
+      "epoch": 8.642201834862385,
+      "grad_norm": 0.393798291683197,
+      "learning_rate": 2.9698818860002797e-05,
+      "loss": 0.0856,
+      "num_input_tokens_seen": 19520352,
+      "step": 475
+    },
+    {
+      "epoch": 8.660550458715596,
+      "grad_norm": 0.379146933555603,
+      "learning_rate": 2.9627373076117863e-05,
+      "loss": 0.0709,
+      "num_input_tokens_seen": 19570472,
+      "step": 476
+    },
+    {
+      "epoch": 8.678899082568808,
+      "grad_norm": 0.4116472005844116,
+      "learning_rate": 2.9555888137303695e-05,
+      "loss": 0.0774,
+      "num_input_tokens_seen": 19600648,
+      "step": 477
+    },
+    {
+      "epoch": 8.697247706422019,
+      "grad_norm": 0.3610764741897583,
+      "learning_rate": 2.9484364648436437e-05,
+      "loss": 0.0758,
+      "num_input_tokens_seen": 19649200,
+      "step": 478
+    },
+    {
+      "epoch": 8.715596330275229,
+      "grad_norm": 0.3889961540699005,
+      "learning_rate": 2.941280321471848e-05,
+      "loss": 0.0833,
+      "num_input_tokens_seen": 19680328,
+      "step": 479
+    },
+    {
+      "epoch": 8.73394495412844,
+      "grad_norm": 0.3711479902267456,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 0.0763,
+      "num_input_tokens_seen": 19728296,
+      "step": 480
+    },
+    {
+      "epoch": 8.752293577981652,
+      "grad_norm": 0.5468257665634155,
+      "learning_rate": 2.9269568935140174e-05,
+      "loss": 0.1141,
+      "num_input_tokens_seen": 19764920,
+      "step": 481
+    },
+    {
+      "epoch": 8.770642201834862,
+      "grad_norm": 0.4884963631629944,
+      "learning_rate": 2.9197897301269435e-05,
+      "loss": 0.137,
+      "num_input_tokens_seen": 19806568,
+      "step": 482
+    },
+    {
+      "epoch": 8.788990825688073,
+      "grad_norm": 0.28893986344337463,
+      "learning_rate": 2.9126190146516942e-05,
+      "loss": 0.059,
+      "num_input_tokens_seen": 19851352,
+      "step": 483
+    },
+    {
+      "epoch": 8.807339449541285,
+      "grad_norm": 0.3916391432285309,
+      "learning_rate": 2.905444807763919e-05,
+      "loss": 0.0993,
+      "num_input_tokens_seen": 19897640,
+      "step": 484
+    },
+    {
+      "epoch": 8.825688073394495,
+      "grad_norm": 0.3886198103427887,
+      "learning_rate": 2.898267170168807e-05,
+      "loss": 0.0721,
+      "num_input_tokens_seen": 19936472,
+      "step": 485
+    },
+    {
+      "epoch": 8.844036697247706,
+      "grad_norm": 0.4273735582828522,
+      "learning_rate": 2.8910861626005776e-05,
+      "loss": 0.0875,
+      "num_input_tokens_seen": 19967128,
+      "step": 486
+    },
+    {
+      "epoch": 8.862385321100918,
+      "grad_norm": 0.3653796315193176,
+      "learning_rate": 2.8839018458219653e-05,
+      "loss": 0.0784,
+      "num_input_tokens_seen": 20010472,
+      "step": 487
+    },
+    {
+      "epoch": 8.880733944954128,
+      "grad_norm": 0.4664580225944519,
+      "learning_rate": 2.876714280623708e-05,
+      "loss": 0.0949,
+      "num_input_tokens_seen": 20056984,
+      "step": 488
+    },
+    {
+      "epoch": 8.899082568807339,
+      "grad_norm": 0.37970954179763794,
+      "learning_rate": 2.8695235278240272e-05,
+      "loss": 0.0718,
+      "num_input_tokens_seen": 20105712,
+      "step": 489
+    },
+    {
+      "epoch": 8.917431192660551,
+      "grad_norm": 0.5023646354675293,
+      "learning_rate": 2.8623296482681166e-05,
+      "loss": 0.0904,
+      "num_input_tokens_seen": 20147136,
+      "step": 490
+    },
+    {
+      "epoch": 8.935779816513762,
+      "grad_norm": 0.4095667004585266,
+      "learning_rate": 2.8551327028276314e-05,
+      "loss": 0.0652,
+      "num_input_tokens_seen": 20190008,
+      "step": 491
+    },
+    {
+      "epoch": 8.954128440366972,
+      "grad_norm": 0.37509602308273315,
+      "learning_rate": 2.8479327524001636e-05,
+      "loss": 0.1347,
+      "num_input_tokens_seen": 20235552,
+      "step": 492
+    },
+    {
+      "epoch": 8.972477064220184,
+      "grad_norm": 0.3403775095939636,
+      "learning_rate": 2.8407298579087365e-05,
+      "loss": 0.0642,
+      "num_input_tokens_seen": 20278504,
+      "step": 493
+    },
+    {
+      "epoch": 8.990825688073395,
+      "grad_norm": 0.3543892204761505,
+      "learning_rate": 2.833524080301282e-05,
+      "loss": 0.0689,
+      "num_input_tokens_seen": 20314024,
+      "step": 494
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.5081737637519836,
+      "learning_rate": 2.8263154805501297e-05,
+      "loss": 0.0735,
+      "num_input_tokens_seen": 20331792,
+      "step": 495
+    },
+    {
+      "epoch": 9.01834862385321,
+      "grad_norm": 0.39798879623413086,
+      "learning_rate": 2.8191041196514873e-05,
+      "loss": 0.0695,
+      "num_input_tokens_seen": 20359504,
+      "step": 496
+    },
+    {
+      "epoch": 9.036697247706423,
+      "grad_norm": 0.4060456454753876,
+      "learning_rate": 2.8118900586249263e-05,
+      "loss": 0.1075,
+      "num_input_tokens_seen": 20393544,
+      "step": 497
+    },
+    {
+      "epoch": 9.055045871559633,
+      "grad_norm": 0.3525485694408417,
+      "learning_rate": 2.8046733585128687e-05,
+      "loss": 0.0658,
+      "num_input_tokens_seen": 20419560,
+      "step": 498
+    },
+    {
+      "epoch": 9.073394495412844,
+      "grad_norm": 0.3256167471408844,
+      "learning_rate": 2.7974540803800637e-05,
+      "loss": 0.0581,
+      "num_input_tokens_seen": 20462936,
+      "step": 499
+    },
+    {
+      "epoch": 9.091743119266056,
+      "grad_norm": 0.5703425407409668,
+      "learning_rate": 2.7902322853130757e-05,
+      "loss": 0.0845,
+      "num_input_tokens_seen": 20494032,
+      "step": 500
+    },
+    {
+      "epoch": 9.110091743119266,
+      "grad_norm": 0.415742963552475,
+      "learning_rate": 2.7830080344197674e-05,
+      "loss": 0.0543,
+      "num_input_tokens_seen": 20553072,
+      "step": 501
+    },
+    {
+      "epoch": 9.128440366972477,
+      "grad_norm": 0.48353156447410583,
+      "learning_rate": 2.7757813888287798e-05,
+      "loss": 0.2288,
+      "num_input_tokens_seen": 20605784,
+      "step": 502
+    },
+    {
+      "epoch": 9.146788990825687,
+      "grad_norm": 0.38978058099746704,
+      "learning_rate": 2.7685524096890185e-05,
+      "loss": 0.0636,
+      "num_input_tokens_seen": 20651856,
+      "step": 503
+    },
+    {
+      "epoch": 9.1651376146789,
+      "grad_norm": 0.3198350965976715,
+      "learning_rate": 2.761321158169134e-05,
+      "loss": 0.0524,
+      "num_input_tokens_seen": 20696944,
+      "step": 504
+    },
+    {
+      "epoch": 9.18348623853211,
+      "grad_norm": 0.39467597007751465,
+      "learning_rate": 2.7540876954570048e-05,
+      "loss": 0.0652,
+      "num_input_tokens_seen": 20728072,
+      "step": 505
+    },
+    {
+      "epoch": 9.20183486238532,
+      "grad_norm": 0.29402676224708557,
+      "learning_rate": 2.7468520827592197e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 20765264,
+      "step": 506
+    },
+    {
+      "epoch": 9.220183486238533,
+      "grad_norm": 0.432449609041214,
+      "learning_rate": 2.7396143813005602e-05,
+      "loss": 0.0763,
+      "num_input_tokens_seen": 20803144,
+      "step": 507
+    },
+    {
+      "epoch": 9.238532110091743,
+      "grad_norm": 0.3020356595516205,
+      "learning_rate": 2.732374652323481e-05,
+      "loss": 0.0495,
+      "num_input_tokens_seen": 20857912,
+      "step": 508
+    },
+    {
+      "epoch": 9.256880733944953,
+      "grad_norm": 0.33522287011146545,
+      "learning_rate": 2.7251329570875934e-05,
+      "loss": 0.0518,
+      "num_input_tokens_seen": 20901536,
+      "step": 509
+    },
+    {
+      "epoch": 9.275229357798166,
+      "grad_norm": 0.3918211758136749,
+      "learning_rate": 2.717889356869146e-05,
+      "loss": 0.0605,
+      "num_input_tokens_seen": 20944592,
+      "step": 510
+    },
+    {
+      "epoch": 9.293577981651376,
+      "grad_norm": 0.32528579235076904,
+      "learning_rate": 2.7106439129605072e-05,
+      "loss": 0.056,
+      "num_input_tokens_seen": 20981640,
+      "step": 511
+    },
+    {
+      "epoch": 9.311926605504587,
+      "grad_norm": 0.45761916041374207,
+      "learning_rate": 2.7033966866696457e-05,
+      "loss": 0.0661,
+      "num_input_tokens_seen": 21011280,
+      "step": 512
+    },
+    {
+      "epoch": 9.330275229357799,
+      "grad_norm": 0.39354491233825684,
+      "learning_rate": 2.6961477393196126e-05,
+      "loss": 0.0735,
+      "num_input_tokens_seen": 21056896,
+      "step": 513
+    },
+    {
+      "epoch": 9.34862385321101,
+      "grad_norm": 0.37278589606285095,
+      "learning_rate": 2.6888971322480205e-05,
+      "loss": 0.0541,
+      "num_input_tokens_seen": 21096440,
+      "step": 514
+    },
+    {
+      "epoch": 9.36697247706422,
+      "grad_norm": 0.547535240650177,
+      "learning_rate": 2.681644926806527e-05,
+      "loss": 0.1135,
+      "num_input_tokens_seen": 21136776,
+      "step": 515
+    },
+    {
+      "epoch": 9.385321100917432,
+      "grad_norm": 0.3297383785247803,
+      "learning_rate": 2.674391184360313e-05,
+      "loss": 0.0572,
+      "num_input_tokens_seen": 21177016,
+      "step": 516
+    },
+    {
+      "epoch": 9.403669724770642,
+      "grad_norm": 0.34031882882118225,
+      "learning_rate": 2.6671359662875684e-05,
+      "loss": 0.0465,
+      "num_input_tokens_seen": 21216216,
+      "step": 517
+    },
+    {
+      "epoch": 9.422018348623853,
+      "grad_norm": 0.36236777901649475,
+      "learning_rate": 2.659879333978964e-05,
+      "loss": 0.0596,
+      "num_input_tokens_seen": 21278344,
+      "step": 518
+    },
+    {
+      "epoch": 9.440366972477065,
+      "grad_norm": 0.3755556344985962,
+      "learning_rate": 2.6526213488371427e-05,
+      "loss": 0.1131,
+      "num_input_tokens_seen": 21330704,
+      "step": 519
+    },
+    {
+      "epoch": 9.458715596330276,
+      "grad_norm": 0.38200709223747253,
+      "learning_rate": 2.6453620722761896e-05,
+      "loss": 0.0587,
+      "num_input_tokens_seen": 21377568,
+      "step": 520
+    },
+    {
+      "epoch": 9.477064220183486,
+      "grad_norm": 0.3465506136417389,
+      "learning_rate": 2.6381015657211215e-05,
+      "loss": 0.0529,
+      "num_input_tokens_seen": 21421808,
+      "step": 521
+    },
+    {
+      "epoch": 9.495412844036696,
+      "grad_norm": 0.3480084538459778,
+      "learning_rate": 2.63083989060736e-05,
+      "loss": 0.0592,
+      "num_input_tokens_seen": 21465904,
+      "step": 522
+    },
+    {
+      "epoch": 9.513761467889909,
+      "grad_norm": 0.3983739912509918,
+      "learning_rate": 2.623577108380215e-05,
+      "loss": 0.0952,
+      "num_input_tokens_seen": 21507888,
+      "step": 523
+    },
+    {
+      "epoch": 9.53211009174312,
+      "grad_norm": 0.43550482392311096,
+      "learning_rate": 2.6163132804943675e-05,
+      "loss": 0.0649,
+      "num_input_tokens_seen": 21541544,
+      "step": 524
+    },
+    {
+      "epoch": 9.55045871559633,
+      "grad_norm": 0.43490955233573914,
+      "learning_rate": 2.6090484684133404e-05,
+      "loss": 0.0557,
+      "num_input_tokens_seen": 21576072,
+      "step": 525
+    },
+    {
+      "epoch": 9.568807339449542,
+      "grad_norm": 0.29154589772224426,
+      "learning_rate": 2.60178273360899e-05,
+      "loss": 0.1018,
+      "num_input_tokens_seen": 21623464,
+      "step": 526
+    },
+    {
+      "epoch": 9.587155963302752,
+      "grad_norm": 0.41565775871276855,
+      "learning_rate": 2.5945161375609778e-05,
+      "loss": 0.0629,
+      "num_input_tokens_seen": 21656096,
+      "step": 527
+    },
+    {
+      "epoch": 9.605504587155963,
+      "grad_norm": 0.3849506080150604,
+      "learning_rate": 2.587248741756253e-05,
+      "loss": 0.0725,
+      "num_input_tokens_seen": 21685872,
+      "step": 528
+    },
+    {
+      "epoch": 9.623853211009175,
+      "grad_norm": 0.425530344247818,
+      "learning_rate": 2.5799806076885326e-05,
+      "loss": 0.0605,
+      "num_input_tokens_seen": 21740704,
+      "step": 529
+    },
+    {
+      "epoch": 9.642201834862385,
+      "grad_norm": 0.47837600111961365,
+      "learning_rate": 2.5727117968577784e-05,
+      "loss": 0.0821,
+      "num_input_tokens_seen": 21775920,
+      "step": 530
+    },
+    {
+      "epoch": 9.660550458715596,
+      "grad_norm": 0.404117226600647,
+      "learning_rate": 2.5654423707696833e-05,
+      "loss": 0.0622,
+      "num_input_tokens_seen": 21822360,
+      "step": 531
+    },
+    {
+      "epoch": 9.678899082568808,
+      "grad_norm": 0.3648509979248047,
+      "learning_rate": 2.5581723909351406e-05,
+      "loss": 0.066,
+      "num_input_tokens_seen": 21872168,
+      "step": 532
+    },
+    {
+      "epoch": 9.697247706422019,
+      "grad_norm": 0.3271898329257965,
+      "learning_rate": 2.5509019188697343e-05,
+      "loss": 0.0555,
+      "num_input_tokens_seen": 21914080,
+      "step": 533
+    },
+    {
+      "epoch": 9.715596330275229,
+      "grad_norm": 0.35200148820877075,
+      "learning_rate": 2.5436310160932092e-05,
+      "loss": 0.0605,
+      "num_input_tokens_seen": 21959648,
+      "step": 534
+    },
+    {
+      "epoch": 9.73394495412844,
+      "grad_norm": 0.35883811116218567,
+      "learning_rate": 2.536359744128957e-05,
+      "loss": 0.0622,
+      "num_input_tokens_seen": 22000280,
+      "step": 535
+    },
+    {
+      "epoch": 9.752293577981652,
+      "grad_norm": 0.3236100673675537,
+      "learning_rate": 2.5290881645034932e-05,
+      "loss": 0.1505,
+      "num_input_tokens_seen": 22058744,
+      "step": 536
+    },
+    {
+      "epoch": 9.770642201834862,
+      "grad_norm": 0.39779284596443176,
+      "learning_rate": 2.521816338745935e-05,
+      "loss": 0.0939,
+      "num_input_tokens_seen": 22102616,
+      "step": 537
+    },
+    {
+      "epoch": 9.788990825688073,
+      "grad_norm": 0.3473663926124573,
+      "learning_rate": 2.5145443283874848e-05,
+      "loss": 0.0507,
+      "num_input_tokens_seen": 22141096,
+      "step": 538
+    },
+    {
+      "epoch": 9.807339449541285,
+      "grad_norm": 0.3695826530456543,
+      "learning_rate": 2.5072721949609053e-05,
+      "loss": 0.0659,
+      "num_input_tokens_seen": 22179672,
+      "step": 539
+    },
+    {
+      "epoch": 9.825688073394495,
+      "grad_norm": 0.41754114627838135,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0683,
+      "num_input_tokens_seen": 22214744,
+      "step": 540
+    },
+    {
+      "epoch": 9.844036697247706,
+      "grad_norm": 0.397276371717453,
+      "learning_rate": 2.4927278050390956e-05,
+      "loss": 0.0633,
+      "num_input_tokens_seen": 22256776,
+      "step": 541
+    },
+    {
+      "epoch": 9.862385321100918,
+      "grad_norm": 0.569098711013794,
+      "learning_rate": 2.485455671612515e-05,
+      "loss": 0.1099,
+      "num_input_tokens_seen": 22300168,
+      "step": 542
+    },
+    {
+      "epoch": 9.880733944954128,
+      "grad_norm": 0.3297954797744751,
+      "learning_rate": 2.4781836612540657e-05,
+      "loss": 0.0571,
+      "num_input_tokens_seen": 22345376,
+      "step": 543
+    },
+    {
+      "epoch": 9.899082568807339,
+      "grad_norm": 0.38287127017974854,
+      "learning_rate": 2.470911835496508e-05,
+      "loss": 0.0552,
+      "num_input_tokens_seen": 22374072,
+      "step": 544
+    },
+    {
+      "epoch": 9.917431192660551,
+      "grad_norm": 0.3791472613811493,
+      "learning_rate": 2.4636402558710432e-05,
+      "loss": 0.1122,
+      "num_input_tokens_seen": 22415592,
+      "step": 545
+    },
+    {
+      "epoch": 9.935779816513762,
+      "grad_norm": 0.3753291368484497,
+      "learning_rate": 2.4563689839067913e-05,
+      "loss": 0.0643,
+      "num_input_tokens_seen": 22458208,
+      "step": 546
+    },
+    {
+      "epoch": 9.954128440366972,
+      "grad_norm": 0.3445354104042053,
+      "learning_rate": 2.4490980811302656e-05,
+      "loss": 0.0535,
+      "num_input_tokens_seen": 22497376,
+      "step": 547
+    },
+    {
+      "epoch": 9.972477064220184,
+      "grad_norm": 0.46824514865875244,
+      "learning_rate": 2.4418276090648596e-05,
+      "loss": 0.1055,
+      "num_input_tokens_seen": 22534016,
+      "step": 548
+    },
+    {
+      "epoch": 9.990825688073395,
+      "grad_norm": 0.3591933846473694,
+      "learning_rate": 2.4345576292303176e-05,
+      "loss": 0.0527,
+      "num_input_tokens_seen": 22573896,
+      "step": 549
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.35627591609954834,
+      "learning_rate": 2.4272882031422215e-05,
+      "loss": 0.0589,
+      "num_input_tokens_seen": 22590880,
+      "step": 550
+    },
+    {
+      "epoch": 10.01834862385321,
+      "grad_norm": 0.36426064372062683,
+      "learning_rate": 2.4200193923114683e-05,
+      "loss": 0.0605,
+      "num_input_tokens_seen": 22629088,
+      "step": 551
+    },
+    {
+      "epoch": 10.036697247706423,
+      "grad_norm": 0.39051762223243713,
+      "learning_rate": 2.4127512582437485e-05,
+      "loss": 0.0537,
+      "num_input_tokens_seen": 22668208,
+      "step": 552
+    },
+    {
+      "epoch": 10.055045871559633,
+      "grad_norm": 0.2818565368652344,
+      "learning_rate": 2.405483862439023e-05,
+      "loss": 0.0458,
+      "num_input_tokens_seen": 22706992,
+      "step": 553
+    },
+    {
+      "epoch": 10.073394495412844,
+      "grad_norm": 0.3922857344150543,
+      "learning_rate": 2.3982172663910108e-05,
+      "loss": 0.057,
+      "num_input_tokens_seen": 22740280,
+      "step": 554
+    },
+    {
+      "epoch": 10.091743119266056,
+      "grad_norm": 0.3783634305000305,
+      "learning_rate": 2.3909515315866605e-05,
+      "loss": 0.1281,
+      "num_input_tokens_seen": 22779464,
+      "step": 555
+    },
+    {
+      "epoch": 10.110091743119266,
+      "grad_norm": 0.3341287076473236,
+      "learning_rate": 2.3836867195056335e-05,
+      "loss": 0.0505,
+      "num_input_tokens_seen": 22819232,
+      "step": 556
+    },
+    {
+      "epoch": 10.128440366972477,
+      "grad_norm": 0.353397399187088,
+      "learning_rate": 2.3764228916197855e-05,
+      "loss": 0.0533,
+      "num_input_tokens_seen": 22860296,
+      "step": 557
+    },
+    {
+      "epoch": 10.146788990825687,
+      "grad_norm": 0.5371742248535156,
+      "learning_rate": 2.3691601093926404e-05,
+      "loss": 0.0657,
+      "num_input_tokens_seen": 22895040,
+      "step": 558
+    },
+    {
+      "epoch": 10.1651376146789,
+      "grad_norm": 0.2916604280471802,
+      "learning_rate": 2.361898434278879e-05,
+      "loss": 0.0448,
+      "num_input_tokens_seen": 22939848,
+      "step": 559
+    },
+    {
+      "epoch": 10.18348623853211,
+      "grad_norm": 0.3729498088359833,
+      "learning_rate": 2.3546379277238107e-05,
+      "loss": 0.0511,
+      "num_input_tokens_seen": 22990280,
+      "step": 560
+    },
+    {
+      "epoch": 10.20183486238532,
+      "grad_norm": 0.39111775159835815,
+      "learning_rate": 2.3473786511628575e-05,
+      "loss": 0.0557,
+      "num_input_tokens_seen": 23034808,
+      "step": 561
+    },
+    {
+      "epoch": 10.220183486238533,
+      "grad_norm": 0.3621370196342468,
+      "learning_rate": 2.3401206660210363e-05,
+      "loss": 0.0481,
+      "num_input_tokens_seen": 23082600,
+      "step": 562
+    },
+    {
+      "epoch": 10.238532110091743,
+      "grad_norm": 0.4028044641017914,
+      "learning_rate": 2.3328640337124326e-05,
+      "loss": 0.0408,
+      "num_input_tokens_seen": 23125248,
+      "step": 563
+    },
+    {
+      "epoch": 10.256880733944953,
+      "grad_norm": 0.7029250264167786,
+      "learning_rate": 2.3256088156396868e-05,
+      "loss": 0.1393,
+      "num_input_tokens_seen": 23169104,
+      "step": 564
+    },
+    {
+      "epoch": 10.275229357798166,
+      "grad_norm": 0.34606480598449707,
+      "learning_rate": 2.3183550731934735e-05,
+      "loss": 0.0445,
+      "num_input_tokens_seen": 23213000,
+      "step": 565
+    },
+    {
+      "epoch": 10.293577981651376,
+      "grad_norm": 0.3310895562171936,
+      "learning_rate": 2.3111028677519804e-05,
+      "loss": 0.0491,
+      "num_input_tokens_seen": 23252584,
+      "step": 566
+    },
+    {
+      "epoch": 10.311926605504587,
+      "grad_norm": 0.3020009398460388,
+      "learning_rate": 2.303852260680388e-05,
+      "loss": 0.0438,
+      "num_input_tokens_seen": 23294424,
+      "step": 567
+    },
+    {
+      "epoch": 10.330275229357799,
+      "grad_norm": 0.41868507862091064,
+      "learning_rate": 2.2966033133303545e-05,
+      "loss": 0.0818,
+      "num_input_tokens_seen": 23331696,
+      "step": 568
+    },
+    {
+      "epoch": 10.34862385321101,
+      "grad_norm": 0.4141708016395569,
+      "learning_rate": 2.289356087039493e-05,
+      "loss": 0.0733,
+      "num_input_tokens_seen": 23370592,
+      "step": 569
+    },
+    {
+      "epoch": 10.36697247706422,
+      "grad_norm": 0.39040666818618774,
+      "learning_rate": 2.2821106431308544e-05,
+      "loss": 0.0576,
+      "num_input_tokens_seen": 23406728,
+      "step": 570
+    },
+    {
+      "epoch": 10.385321100917432,
+      "grad_norm": 0.29972052574157715,
+      "learning_rate": 2.274867042912408e-05,
+      "loss": 0.0399,
+      "num_input_tokens_seen": 23456880,
+      "step": 571
+    },
+    {
+      "epoch": 10.403669724770642,
+      "grad_norm": 0.4595007002353668,
+      "learning_rate": 2.2676253476765196e-05,
+      "loss": 0.1006,
+      "num_input_tokens_seen": 23501632,
+      "step": 572
+    },
+    {
+      "epoch": 10.422018348623853,
+      "grad_norm": 0.3600683808326721,
+      "learning_rate": 2.26038561869944e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 23546696,
+      "step": 573
+    },
+    {
+      "epoch": 10.440366972477065,
+      "grad_norm": 0.33539336919784546,
+      "learning_rate": 2.2531479172407805e-05,
+      "loss": 0.038,
+      "num_input_tokens_seen": 23590976,
+      "step": 574
+    },
+    {
+      "epoch": 10.458715596330276,
+      "grad_norm": 0.37528175115585327,
+      "learning_rate": 2.2459123045429954e-05,
+      "loss": 0.0475,
+      "num_input_tokens_seen": 23632096,
+      "step": 575
+    },
+    {
+      "epoch": 10.477064220183486,
+      "grad_norm": 0.4022611379623413,
+      "learning_rate": 2.238678841830867e-05,
+      "loss": 0.0385,
+      "num_input_tokens_seen": 23661968,
+      "step": 576
+    },
+    {
+      "epoch": 10.495412844036696,
+      "grad_norm": 0.5055795907974243,
+      "learning_rate": 2.2314475903109825e-05,
+      "loss": 0.0602,
+      "num_input_tokens_seen": 23709568,
+      "step": 577
+    },
+    {
+      "epoch": 10.513761467889909,
+      "grad_norm": 0.3705989122390747,
+      "learning_rate": 2.2242186111712208e-05,
+      "loss": 0.0812,
+      "num_input_tokens_seen": 23749216,
+      "step": 578
+    },
+    {
+      "epoch": 10.53211009174312,
+      "grad_norm": 0.46322381496429443,
+      "learning_rate": 2.2169919655802335e-05,
+      "loss": 0.1103,
+      "num_input_tokens_seen": 23786848,
+      "step": 579
+    },
+    {
+      "epoch": 10.55045871559633,
+      "grad_norm": 0.40934255719184875,
+      "learning_rate": 2.2097677146869242e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 23827336,
+      "step": 580
+    },
+    {
+      "epoch": 10.568807339449542,
+      "grad_norm": 0.7357956767082214,
+      "learning_rate": 2.202545919619937e-05,
+      "loss": 0.1118,
+      "num_input_tokens_seen": 23880272,
+      "step": 581
+    },
+    {
+      "epoch": 10.587155963302752,
+      "grad_norm": 0.4258132576942444,
+      "learning_rate": 2.195326641487132e-05,
+      "loss": 0.0512,
+      "num_input_tokens_seen": 23923208,
+      "step": 582
+    },
+    {
+      "epoch": 10.605504587155963,
+      "grad_norm": 0.3786598742008209,
+      "learning_rate": 2.1881099413750733e-05,
+      "loss": 0.0809,
+      "num_input_tokens_seen": 23961424,
+      "step": 583
+    },
+    {
+      "epoch": 10.623853211009175,
+      "grad_norm": 0.37131381034851074,
+      "learning_rate": 2.1808958803485136e-05,
+      "loss": 0.0446,
+      "num_input_tokens_seen": 24000184,
+      "step": 584
+    },
+    {
+      "epoch": 10.642201834862385,
+      "grad_norm": 0.3253346383571625,
+      "learning_rate": 2.173684519449872e-05,
+      "loss": 0.0472,
+      "num_input_tokens_seen": 24049488,
+      "step": 585
+    },
+    {
+      "epoch": 10.660550458715596,
+      "grad_norm": 0.4104836881160736,
+      "learning_rate": 2.1664759196987182e-05,
+      "loss": 0.0522,
+      "num_input_tokens_seen": 24085104,
+      "step": 586
+    },
+    {
+      "epoch": 10.678899082568808,
+      "grad_norm": 0.5492331385612488,
+      "learning_rate": 2.1592701420912644e-05,
+      "loss": 0.1182,
+      "num_input_tokens_seen": 24124768,
+      "step": 587
+    },
+    {
+      "epoch": 10.697247706422019,
+      "grad_norm": 0.315057635307312,
+      "learning_rate": 2.1520672475998373e-05,
+      "loss": 0.0452,
+      "num_input_tokens_seen": 24167872,
+      "step": 588
+    },
+    {
+      "epoch": 10.715596330275229,
+      "grad_norm": 0.3825458586215973,
+      "learning_rate": 2.144867297172369e-05,
+      "loss": 0.0586,
+      "num_input_tokens_seen": 24198824,
+      "step": 589
+    },
+    {
+      "epoch": 10.73394495412844,
+      "grad_norm": 0.3713642656803131,
+      "learning_rate": 2.1376703517318837e-05,
+      "loss": 0.0547,
+      "num_input_tokens_seen": 24249320,
+      "step": 590
+    },
+    {
+      "epoch": 10.752293577981652,
+      "grad_norm": 0.4260292947292328,
+      "learning_rate": 2.1304764721759733e-05,
+      "loss": 0.0562,
+      "num_input_tokens_seen": 24292592,
+      "step": 591
+    },
+    {
+      "epoch": 10.770642201834862,
+      "grad_norm": 0.4098452627658844,
+      "learning_rate": 2.1232857193762924e-05,
+      "loss": 0.056,
+      "num_input_tokens_seen": 24331648,
+      "step": 592
+    },
+    {
+      "epoch": 10.788990825688073,
+      "grad_norm": 0.4303808808326721,
+      "learning_rate": 2.116098154178035e-05,
+      "loss": 0.0538,
+      "num_input_tokens_seen": 24386904,
+      "step": 593
+    },
+    {
+      "epoch": 10.807339449541285,
+      "grad_norm": 0.41580283641815186,
+      "learning_rate": 2.1089138373994223e-05,
+      "loss": 0.0924,
+      "num_input_tokens_seen": 24435360,
+      "step": 594
+    },
+    {
+      "epoch": 10.825688073394495,
+      "grad_norm": 0.3460528254508972,
+      "learning_rate": 2.101732829831194e-05,
+      "loss": 0.0468,
+      "num_input_tokens_seen": 24465400,
+      "step": 595
+    },
+    {
+      "epoch": 10.844036697247706,
+      "grad_norm": 0.4775545299053192,
+      "learning_rate": 2.0945551922360818e-05,
+      "loss": 0.057,
+      "num_input_tokens_seen": 24497984,
+      "step": 596
+    },
+    {
+      "epoch": 10.862385321100918,
+      "grad_norm": 0.36002540588378906,
+      "learning_rate": 2.087380985348306e-05,
+      "loss": 0.0472,
+      "num_input_tokens_seen": 24534576,
+      "step": 597
+    },
+    {
+      "epoch": 10.880733944954128,
+      "grad_norm": 0.38081276416778564,
+      "learning_rate": 2.0802102698730574e-05,
+      "loss": 0.0592,
+      "num_input_tokens_seen": 24572160,
+      "step": 598
+    },
+    {
+      "epoch": 10.899082568807339,
+      "grad_norm": 0.3270852863788605,
+      "learning_rate": 2.0730431064859836e-05,
+      "loss": 0.0392,
+      "num_input_tokens_seen": 24619744,
+      "step": 599
+    },
+    {
+      "epoch": 10.917431192660551,
+      "grad_norm": 0.36131611466407776,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 24671048,
+      "step": 600
+    },
+    {
+      "epoch": 10.935779816513762,
+      "grad_norm": 0.40473103523254395,
+      "learning_rate": 2.0587196785281525e-05,
+      "loss": 0.0512,
+      "num_input_tokens_seen": 24710152,
+      "step": 601
+    },
+    {
+      "epoch": 10.954128440366972,
+      "grad_norm": 0.35128137469291687,
+      "learning_rate": 2.0515635351563565e-05,
+      "loss": 0.0466,
+      "num_input_tokens_seen": 24744584,
+      "step": 602
+    },
+    {
+      "epoch": 10.972477064220184,
+      "grad_norm": 0.6386734843254089,
+      "learning_rate": 2.0444111862696314e-05,
+      "loss": 0.0911,
+      "num_input_tokens_seen": 24788584,
+      "step": 603
+    },
+    {
+      "epoch": 10.990825688073395,
+      "grad_norm": 0.4696199297904968,
+      "learning_rate": 2.037262692388214e-05,
+      "loss": 0.0516,
+      "num_input_tokens_seen": 24823152,
+      "step": 604
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.4481979310512543,
+      "learning_rate": 2.0301181139997205e-05,
+      "loss": 0.1007,
+      "num_input_tokens_seen": 24849968,
+      "step": 605
+    },
+    {
+      "epoch": 11.01834862385321,
+      "grad_norm": 0.3045099079608917,
+      "learning_rate": 2.022977511558638e-05,
+      "loss": 0.0425,
+      "num_input_tokens_seen": 24892784,
+      "step": 606
+    },
+    {
+      "epoch": 11.036697247706423,
+      "grad_norm": 0.2954024374485016,
+      "learning_rate": 2.0158409454858103e-05,
+      "loss": 0.0887,
+      "num_input_tokens_seen": 24937192,
+      "step": 607
+    },
+    {
+      "epoch": 11.055045871559633,
+      "grad_norm": 0.302165150642395,
+      "learning_rate": 2.0087084761679245e-05,
+      "loss": 0.0425,
+      "num_input_tokens_seen": 24980608,
+      "step": 608
+    },
+    {
+      "epoch": 11.073394495412844,
+      "grad_norm": 0.36207833886146545,
+      "learning_rate": 2.0015801639570074e-05,
+      "loss": 0.0788,
+      "num_input_tokens_seen": 25029016,
+      "step": 609
+    },
+    {
+      "epoch": 11.091743119266056,
+      "grad_norm": 0.3311103582382202,
+      "learning_rate": 1.9944560691699057e-05,
+      "loss": 0.0387,
+      "num_input_tokens_seen": 25071880,
+      "step": 610
+    },
+    {
+      "epoch": 11.110091743119266,
+      "grad_norm": 0.3376132547855377,
+      "learning_rate": 1.9873362520877813e-05,
+      "loss": 0.0534,
+      "num_input_tokens_seen": 25117112,
+      "step": 611
+    },
+    {
+      "epoch": 11.128440366972477,
+      "grad_norm": 0.3412828743457794,
+      "learning_rate": 1.980220772955602e-05,
+      "loss": 0.0509,
+      "num_input_tokens_seen": 25154456,
+      "step": 612
+    },
+    {
+      "epoch": 11.146788990825687,
+      "grad_norm": 0.40508487820625305,
+      "learning_rate": 1.973109691981627e-05,
+      "loss": 0.0436,
+      "num_input_tokens_seen": 25193264,
+      "step": 613
+    },
+    {
+      "epoch": 11.1651376146789,
+      "grad_norm": 0.3498331606388092,
+      "learning_rate": 1.9660030693369004e-05,
+      "loss": 0.04,
+      "num_input_tokens_seen": 25231416,
+      "step": 614
+    },
+    {
+      "epoch": 11.18348623853211,
+      "grad_norm": 0.36831074953079224,
+      "learning_rate": 1.958900965154743e-05,
+      "loss": 0.0371,
+      "num_input_tokens_seen": 25268136,
+      "step": 615
+    },
+    {
+      "epoch": 11.20183486238532,
+      "grad_norm": 0.4536377787590027,
+      "learning_rate": 1.9518034395302414e-05,
+      "loss": 0.0986,
+      "num_input_tokens_seen": 25316600,
+      "step": 616
+    },
+    {
+      "epoch": 11.220183486238533,
+      "grad_norm": 0.3427221477031708,
+      "learning_rate": 1.9447105525197425e-05,
+      "loss": 0.0391,
+      "num_input_tokens_seen": 25358424,
+      "step": 617
+    },
+    {
+      "epoch": 11.238532110091743,
+      "grad_norm": 0.3086804151535034,
+      "learning_rate": 1.937622364140338e-05,
+      "loss": 0.0378,
+      "num_input_tokens_seen": 25403744,
+      "step": 618
+    },
+    {
+      "epoch": 11.256880733944953,
+      "grad_norm": 0.37512463331222534,
+      "learning_rate": 1.9305389343693664e-05,
+      "loss": 0.0412,
+      "num_input_tokens_seen": 25436496,
+      "step": 619
+    },
+    {
+      "epoch": 11.275229357798166,
+      "grad_norm": 0.48066776990890503,
+      "learning_rate": 1.9234603231438995e-05,
+      "loss": 0.0473,
+      "num_input_tokens_seen": 25476544,
+      "step": 620
+    },
+    {
+      "epoch": 11.293577981651376,
+      "grad_norm": 0.3921424448490143,
+      "learning_rate": 1.9163865903602374e-05,
+      "loss": 0.0874,
+      "num_input_tokens_seen": 25514768,
+      "step": 621
+    },
+    {
+      "epoch": 11.311926605504587,
+      "grad_norm": 0.36180391907691956,
+      "learning_rate": 1.9093177958733966e-05,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 25562136,
+      "step": 622
+    },
+    {
+      "epoch": 11.330275229357799,
+      "grad_norm": 0.3051501512527466,
+      "learning_rate": 1.9022539994966147e-05,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 25599784,
+      "step": 623
+    },
+    {
+      "epoch": 11.34862385321101,
+      "grad_norm": 0.3278719186782837,
+      "learning_rate": 1.895195261000831e-05,
+      "loss": 0.0368,
+      "num_input_tokens_seen": 25633664,
+      "step": 624
+    },
+    {
+      "epoch": 11.36697247706422,
+      "grad_norm": 0.31365832686424255,
+      "learning_rate": 1.8881416401141904e-05,
+      "loss": 0.0817,
+      "num_input_tokens_seen": 25686344,
+      "step": 625
+    },
+    {
+      "epoch": 11.385321100917432,
+      "grad_norm": 0.31113532185554504,
+      "learning_rate": 1.8810931965215356e-05,
+      "loss": 0.042,
+      "num_input_tokens_seen": 25731184,
+      "step": 626
+    },
+    {
+      "epoch": 11.403669724770642,
+      "grad_norm": 0.6552867889404297,
+      "learning_rate": 1.874049989863896e-05,
+      "loss": 0.0803,
+      "num_input_tokens_seen": 25773632,
+      "step": 627
+    },
+    {
+      "epoch": 11.422018348623853,
+      "grad_norm": 0.3581204116344452,
+      "learning_rate": 1.8670120797379958e-05,
+      "loss": 0.043,
+      "num_input_tokens_seen": 25803376,
+      "step": 628
+    },
+    {
+      "epoch": 11.440366972477065,
+      "grad_norm": 0.308876097202301,
+      "learning_rate": 1.859979525695736e-05,
+      "loss": 0.0389,
+      "num_input_tokens_seen": 25855248,
+      "step": 629
+    },
+    {
+      "epoch": 11.458715596330276,
+      "grad_norm": 0.3815799951553345,
+      "learning_rate": 1.852952387243698e-05,
+      "loss": 0.0429,
+      "num_input_tokens_seen": 25885840,
+      "step": 630
+    },
+    {
+      "epoch": 11.477064220183486,
+      "grad_norm": 0.34968090057373047,
+      "learning_rate": 1.8459307238426416e-05,
+      "loss": 0.0466,
+      "num_input_tokens_seen": 25938800,
+      "step": 631
+    },
+    {
+      "epoch": 11.495412844036696,
+      "grad_norm": 0.4210070073604584,
+      "learning_rate": 1.838914594906995e-05,
+      "loss": 0.0543,
+      "num_input_tokens_seen": 25973752,
+      "step": 632
+    },
+    {
+      "epoch": 11.513761467889909,
+      "grad_norm": 0.3628455698490143,
+      "learning_rate": 1.831904059804358e-05,
+      "loss": 0.0408,
+      "num_input_tokens_seen": 26011480,
+      "step": 633
+    },
+    {
+      "epoch": 11.53211009174312,
+      "grad_norm": 0.3586367964744568,
+      "learning_rate": 1.8248991778549984e-05,
+      "loss": 0.0462,
+      "num_input_tokens_seen": 26054744,
+      "step": 634
+    },
+    {
+      "epoch": 11.55045871559633,
+      "grad_norm": 0.44129693508148193,
+      "learning_rate": 1.8179000083313483e-05,
+      "loss": 0.0478,
+      "num_input_tokens_seen": 26083512,
+      "step": 635
+    },
+    {
+      "epoch": 11.568807339449542,
+      "grad_norm": 0.3320532739162445,
+      "learning_rate": 1.8109066104575023e-05,
+      "loss": 0.0376,
+      "num_input_tokens_seen": 26127504,
+      "step": 636
+    },
+    {
+      "epoch": 11.587155963302752,
+      "grad_norm": 0.34192633628845215,
+      "learning_rate": 1.8039190434087212e-05,
+      "loss": 0.0399,
+      "num_input_tokens_seen": 26170568,
+      "step": 637
+    },
+    {
+      "epoch": 11.605504587155963,
+      "grad_norm": 0.5145437121391296,
+      "learning_rate": 1.7969373663109234e-05,
+      "loss": 0.0658,
+      "num_input_tokens_seen": 26221816,
+      "step": 638
+    },
+    {
+      "epoch": 11.623853211009175,
+      "grad_norm": 0.3665778636932373,
+      "learning_rate": 1.7899616382401936e-05,
+      "loss": 0.0384,
+      "num_input_tokens_seen": 26262800,
+      "step": 639
+    },
+    {
+      "epoch": 11.642201834862385,
+      "grad_norm": 0.31278637051582336,
+      "learning_rate": 1.7829919182222752e-05,
+      "loss": 0.0357,
+      "num_input_tokens_seen": 26311200,
+      "step": 640
+    },
+    {
+      "epoch": 11.660550458715596,
+      "grad_norm": 0.3833770751953125,
+      "learning_rate": 1.776028265232073e-05,
+      "loss": 0.0445,
+      "num_input_tokens_seen": 26359400,
+      "step": 641
+    },
+    {
+      "epoch": 11.678899082568808,
+      "grad_norm": 0.3182731568813324,
+      "learning_rate": 1.7690707381931583e-05,
+      "loss": 0.0352,
+      "num_input_tokens_seen": 26401144,
+      "step": 642
+    },
+    {
+      "epoch": 11.697247706422019,
+      "grad_norm": 0.3680736720561981,
+      "learning_rate": 1.7621193959772657e-05,
+      "loss": 0.0381,
+      "num_input_tokens_seen": 26429056,
+      "step": 643
+    },
+    {
+      "epoch": 11.715596330275229,
+      "grad_norm": 0.30601194500923157,
+      "learning_rate": 1.755174297403795e-05,
+      "loss": 0.0448,
+      "num_input_tokens_seen": 26478272,
+      "step": 644
+    },
+    {
+      "epoch": 11.73394495412844,
+      "grad_norm": 0.35907667875289917,
+      "learning_rate": 1.7482355012393177e-05,
+      "loss": 0.044,
+      "num_input_tokens_seen": 26525592,
+      "step": 645
+    },
+    {
+      "epoch": 11.752293577981652,
+      "grad_norm": 0.33723318576812744,
+      "learning_rate": 1.7413030661970742e-05,
+      "loss": 0.0392,
+      "num_input_tokens_seen": 26565280,
+      "step": 646
+    },
+    {
+      "epoch": 11.770642201834862,
+      "grad_norm": 0.3269751965999603,
+      "learning_rate": 1.73437705093648e-05,
+      "loss": 0.0356,
+      "num_input_tokens_seen": 26601768,
+      "step": 647
+    },
+    {
+      "epoch": 11.788990825688073,
+      "grad_norm": 0.339697927236557,
+      "learning_rate": 1.7274575140626318e-05,
+      "loss": 0.0425,
+      "num_input_tokens_seen": 26634496,
+      "step": 648
+    },
+    {
+      "epoch": 11.807339449541285,
+      "grad_norm": 0.37456247210502625,
+      "learning_rate": 1.720544514125805e-05,
+      "loss": 0.0504,
+      "num_input_tokens_seen": 26679232,
+      "step": 649
+    },
+    {
+      "epoch": 11.825688073394495,
+      "grad_norm": 0.42360228300094604,
+      "learning_rate": 1.7136381096209664e-05,
+      "loss": 0.0566,
+      "num_input_tokens_seen": 26723144,
+      "step": 650
+    },
+    {
+      "epoch": 11.844036697247706,
+      "grad_norm": 0.3331366181373596,
+      "learning_rate": 1.7067383589872703e-05,
+      "loss": 0.0397,
+      "num_input_tokens_seen": 26760448,
+      "step": 651
+    },
+    {
+      "epoch": 11.862385321100918,
+      "grad_norm": 0.30122336745262146,
+      "learning_rate": 1.699845320607571e-05,
+      "loss": 0.0372,
+      "num_input_tokens_seen": 26820856,
+      "step": 652
+    },
+    {
+      "epoch": 11.880733944954128,
+      "grad_norm": 0.8716153502464294,
+      "learning_rate": 1.692959052807928e-05,
+      "loss": 0.0932,
+      "num_input_tokens_seen": 26851736,
+      "step": 653
+    },
+    {
+      "epoch": 11.899082568807339,
+      "grad_norm": 0.3729649484157562,
+      "learning_rate": 1.686079613857109e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 26892976,
+      "step": 654
+    },
+    {
+      "epoch": 11.917431192660551,
+      "grad_norm": 0.4268723130226135,
+      "learning_rate": 1.6792070619660975e-05,
+      "loss": 0.0806,
+      "num_input_tokens_seen": 26933888,
+      "step": 655
+    },
+    {
+      "epoch": 11.935779816513762,
+      "grad_norm": 0.3705332279205322,
+      "learning_rate": 1.672341455287605e-05,
+      "loss": 0.0414,
+      "num_input_tokens_seen": 26970576,
+      "step": 656
+    },
+    {
+      "epoch": 11.954128440366972,
+      "grad_norm": 0.3180471360683441,
+      "learning_rate": 1.665482851915573e-05,
+      "loss": 0.0356,
+      "num_input_tokens_seen": 27008400,
+      "step": 657
+    },
+    {
+      "epoch": 11.972477064220184,
+      "grad_norm": 0.6832396984100342,
+      "learning_rate": 1.658631309884684e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 27043704,
+      "step": 658
+    },
+    {
+      "epoch": 11.990825688073395,
+      "grad_norm": 0.4499323070049286,
+      "learning_rate": 1.6517868871698725e-05,
+      "loss": 0.0473,
+      "num_input_tokens_seen": 27084456,
+      "step": 659
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.4537651240825653,
+      "learning_rate": 1.6449496416858284e-05,
+      "loss": 0.0369,
+      "num_input_tokens_seen": 27109056,
+      "step": 660
+    },
+    {
+      "epoch": 12.01834862385321,
+      "grad_norm": 0.3181232213973999,
+      "learning_rate": 1.6381196312865145e-05,
+      "loss": 0.0767,
+      "num_input_tokens_seen": 27147704,
+      "step": 661
+    },
+    {
+      "epoch": 12.036697247706423,
+      "grad_norm": 0.30549949407577515,
+      "learning_rate": 1.6312969137646716e-05,
+      "loss": 0.0398,
+      "num_input_tokens_seen": 27197376,
+      "step": 662
+    },
+    {
+      "epoch": 12.055045871559633,
+      "grad_norm": 0.29217857122421265,
+      "learning_rate": 1.6244815468513315e-05,
+      "loss": 0.0373,
+      "num_input_tokens_seen": 27244056,
+      "step": 663
+    },
+    {
+      "epoch": 12.073394495412844,
+      "grad_norm": 0.3700779378414154,
+      "learning_rate": 1.617673588215328e-05,
+      "loss": 0.042,
+      "num_input_tokens_seen": 27269576,
+      "step": 664
+    },
+    {
+      "epoch": 12.091743119266056,
+      "grad_norm": 0.3830711245536804,
+      "learning_rate": 1.6108730954628093e-05,
+      "loss": 0.0317,
+      "num_input_tokens_seen": 27314640,
+      "step": 665
+    },
+    {
+      "epoch": 12.110091743119266,
+      "grad_norm": 0.332509845495224,
+      "learning_rate": 1.6040801261367493e-05,
+      "loss": 0.0382,
+      "num_input_tokens_seen": 27358840,
+      "step": 666
+    },
+    {
+      "epoch": 12.128440366972477,
+      "grad_norm": 0.35123559832572937,
+      "learning_rate": 1.5972947377164645e-05,
+      "loss": 0.0378,
+      "num_input_tokens_seen": 27399112,
+      "step": 667
+    },
+    {
+      "epoch": 12.146788990825687,
+      "grad_norm": 0.3346501588821411,
+      "learning_rate": 1.5905169876171223e-05,
+      "loss": 0.0322,
+      "num_input_tokens_seen": 27434064,
+      "step": 668
+    },
+    {
+      "epoch": 12.1651376146789,
+      "grad_norm": 0.38763749599456787,
+      "learning_rate": 1.583746933189257e-05,
+      "loss": 0.0421,
+      "num_input_tokens_seen": 27468528,
+      "step": 669
+    },
+    {
+      "epoch": 12.18348623853211,
+      "grad_norm": 0.3860892653465271,
+      "learning_rate": 1.5769846317182893e-05,
+      "loss": 0.0682,
+      "num_input_tokens_seen": 27508512,
+      "step": 670
+    },
+    {
+      "epoch": 12.20183486238532,
+      "grad_norm": 0.31347358226776123,
+      "learning_rate": 1.570230140424033e-05,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 27555912,
+      "step": 671
+    },
+    {
+      "epoch": 12.220183486238533,
+      "grad_norm": 0.3007822632789612,
+      "learning_rate": 1.56348351646022e-05,
+      "loss": 0.0346,
+      "num_input_tokens_seen": 27597608,
+      "step": 672
+    },
+    {
+      "epoch": 12.238532110091743,
+      "grad_norm": 0.38240882754325867,
+      "learning_rate": 1.556744816914008e-05,
+      "loss": 0.037,
+      "num_input_tokens_seen": 27633496,
+      "step": 673
+    },
+    {
+      "epoch": 12.256880733944953,
+      "grad_norm": 0.2740360498428345,
+      "learning_rate": 1.550014098805503e-05,
+      "loss": 0.0286,
+      "num_input_tokens_seen": 27680040,
+      "step": 674
+    },
+    {
+      "epoch": 12.275229357798166,
+      "grad_norm": 0.4563816487789154,
+      "learning_rate": 1.5432914190872757e-05,
+      "loss": 0.0308,
+      "num_input_tokens_seen": 27721872,
+      "step": 675
+    },
+    {
+      "epoch": 12.293577981651376,
+      "grad_norm": 0.27564939856529236,
+      "learning_rate": 1.5365768346438797e-05,
+      "loss": 0.0294,
+      "num_input_tokens_seen": 27762968,
+      "step": 676
+    },
+    {
+      "epoch": 12.311926605504587,
+      "grad_norm": 0.28814566135406494,
+      "learning_rate": 1.529870402291368e-05,
+      "loss": 0.0283,
+      "num_input_tokens_seen": 27809496,
+      "step": 677
+    },
+    {
+      "epoch": 12.330275229357799,
+      "grad_norm": 0.29550161957740784,
+      "learning_rate": 1.523172178776816e-05,
+      "loss": 0.027,
+      "num_input_tokens_seen": 27857160,
+      "step": 678
+    },
+    {
+      "epoch": 12.34862385321101,
+      "grad_norm": 0.31906190514564514,
+      "learning_rate": 1.5164822207778379e-05,
+      "loss": 0.0351,
+      "num_input_tokens_seen": 27920616,
+      "step": 679
+    },
+    {
+      "epoch": 12.36697247706422,
+      "grad_norm": 0.4574970602989197,
+      "learning_rate": 1.509800584902108e-05,
+      "loss": 0.0749,
+      "num_input_tokens_seen": 27971920,
+      "step": 680
+    },
+    {
+      "epoch": 12.385321100917432,
+      "grad_norm": 0.4354063868522644,
+      "learning_rate": 1.5031273276868845e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 28008960,
+      "step": 681
+    },
+    {
+      "epoch": 12.403669724770642,
+      "grad_norm": 0.32136741280555725,
+      "learning_rate": 1.4964625055985265e-05,
+      "loss": 0.0319,
+      "num_input_tokens_seen": 28055416,
+      "step": 682
+    },
+    {
+      "epoch": 12.422018348623853,
+      "grad_norm": 0.3991897702217102,
+      "learning_rate": 1.4898061750320212e-05,
+      "loss": 0.0403,
+      "num_input_tokens_seen": 28093136,
+      "step": 683
+    },
+    {
+      "epoch": 12.440366972477065,
+      "grad_norm": 0.29797106981277466,
+      "learning_rate": 1.4831583923104999e-05,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 28152744,
+      "step": 684
+    },
+    {
+      "epoch": 12.458715596330276,
+      "grad_norm": 0.8847760558128357,
+      "learning_rate": 1.4765192136847685e-05,
+      "loss": 0.0631,
+      "num_input_tokens_seen": 28205712,
+      "step": 685
+    },
+    {
+      "epoch": 12.477064220183486,
+      "grad_norm": 0.28264713287353516,
+      "learning_rate": 1.4698886953328292e-05,
+      "loss": 0.0304,
+      "num_input_tokens_seen": 28248432,
+      "step": 686
+    },
+    {
+      "epoch": 12.495412844036696,
+      "grad_norm": 0.5237520337104797,
+      "learning_rate": 1.463266893359403e-05,
+      "loss": 0.0612,
+      "num_input_tokens_seen": 28282728,
+      "step": 687
+    },
+    {
+      "epoch": 12.513761467889909,
+      "grad_norm": 0.3867344260215759,
+      "learning_rate": 1.4566538637954554e-05,
+      "loss": 0.0436,
+      "num_input_tokens_seen": 28330200,
+      "step": 688
+    },
+    {
+      "epoch": 12.53211009174312,
+      "grad_norm": 0.34402939677238464,
+      "learning_rate": 1.4500496625977264e-05,
+      "loss": 0.0339,
+      "num_input_tokens_seen": 28368752,
+      "step": 689
+    },
+    {
+      "epoch": 12.55045871559633,
+      "grad_norm": 0.3360813558101654,
+      "learning_rate": 1.443454345648252e-05,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 28405528,
+      "step": 690
+    },
+    {
+      "epoch": 12.568807339449542,
+      "grad_norm": 0.3713105618953705,
+      "learning_rate": 1.436867968753893e-05,
+      "loss": 0.0378,
+      "num_input_tokens_seen": 28451400,
+      "step": 691
+    },
+    {
+      "epoch": 12.587155963302752,
+      "grad_norm": 0.328673392534256,
+      "learning_rate": 1.430290587645865e-05,
+      "loss": 0.0407,
+      "num_input_tokens_seen": 28501600,
+      "step": 692
+    },
+    {
+      "epoch": 12.605504587155963,
+      "grad_norm": 0.37341922521591187,
+      "learning_rate": 1.4237222579792618e-05,
+      "loss": 0.0362,
+      "num_input_tokens_seen": 28532400,
+      "step": 693
+    },
+    {
+      "epoch": 12.623853211009175,
+      "grad_norm": 0.416866660118103,
+      "learning_rate": 1.4171630353325932e-05,
+      "loss": 0.0423,
+      "num_input_tokens_seen": 28568872,
+      "step": 694
+    },
+    {
+      "epoch": 12.642201834862385,
+      "grad_norm": 0.3353414535522461,
+      "learning_rate": 1.4106129752073022e-05,
+      "loss": 0.0365,
+      "num_input_tokens_seen": 28610976,
+      "step": 695
+    },
+    {
+      "epoch": 12.660550458715596,
+      "grad_norm": 0.4057430028915405,
+      "learning_rate": 1.4040721330273062e-05,
+      "loss": 0.083,
+      "num_input_tokens_seen": 28657064,
+      "step": 696
+    },
+    {
+      "epoch": 12.678899082568808,
+      "grad_norm": 0.26201528310775757,
+      "learning_rate": 1.3975405641385252e-05,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 28702912,
+      "step": 697
+    },
+    {
+      "epoch": 12.697247706422019,
+      "grad_norm": 0.3357066214084625,
+      "learning_rate": 1.3910183238084112e-05,
+      "loss": 0.0346,
+      "num_input_tokens_seen": 28754368,
+      "step": 698
+    },
+    {
+      "epoch": 12.715596330275229,
+      "grad_norm": 0.30762168765068054,
+      "learning_rate": 1.3845054672254781e-05,
+      "loss": 0.0289,
+      "num_input_tokens_seen": 28795120,
+      "step": 699
+    },
+    {
+      "epoch": 12.73394495412844,
+      "grad_norm": 0.30095961689949036,
+      "learning_rate": 1.3780020494988446e-05,
+      "loss": 0.0288,
+      "num_input_tokens_seen": 28837920,
+      "step": 700
+    },
+    {
+      "epoch": 12.752293577981652,
+      "grad_norm": 0.4343203008174896,
+      "learning_rate": 1.3715081256577582e-05,
+      "loss": 0.041,
+      "num_input_tokens_seen": 28874200,
+      "step": 701
+    },
+    {
+      "epoch": 12.770642201834862,
+      "grad_norm": 0.37679317593574524,
+      "learning_rate": 1.3650237506511331e-05,
+      "loss": 0.0304,
+      "num_input_tokens_seen": 28911744,
+      "step": 702
+    },
+    {
+      "epoch": 12.788990825688073,
+      "grad_norm": 0.3850303292274475,
+      "learning_rate": 1.3585489793470862e-05,
+      "loss": 0.0292,
+      "num_input_tokens_seen": 28938696,
+      "step": 703
+    },
+    {
+      "epoch": 12.807339449541285,
+      "grad_norm": 0.35162273049354553,
+      "learning_rate": 1.3520838665324703e-05,
+      "loss": 0.0321,
+      "num_input_tokens_seen": 28978360,
+      "step": 704
+    },
+    {
+      "epoch": 12.825688073394495,
+      "grad_norm": 0.35139572620391846,
+      "learning_rate": 1.3456284669124158e-05,
+      "loss": 0.0311,
+      "num_input_tokens_seen": 29008048,
+      "step": 705
+    },
+    {
+      "epoch": 12.844036697247706,
+      "grad_norm": 0.3229581117630005,
+      "learning_rate": 1.3391828351098578e-05,
+      "loss": 0.0307,
+      "num_input_tokens_seen": 29047928,
+      "step": 706
+    },
+    {
+      "epoch": 12.862385321100918,
+      "grad_norm": 0.3116603493690491,
+      "learning_rate": 1.3327470256650848e-05,
+      "loss": 0.033,
+      "num_input_tokens_seen": 29080376,
+      "step": 707
+    },
+    {
+      "epoch": 12.880733944954128,
+      "grad_norm": 0.331969290971756,
+      "learning_rate": 1.3263210930352737e-05,
+      "loss": 0.0299,
+      "num_input_tokens_seen": 29113856,
+      "step": 708
+    },
+    {
+      "epoch": 12.899082568807339,
+      "grad_norm": 0.32986903190612793,
+      "learning_rate": 1.3199050915940225e-05,
+      "loss": 0.0373,
+      "num_input_tokens_seen": 29162904,
+      "step": 709
+    },
+    {
+      "epoch": 12.917431192660551,
+      "grad_norm": 0.3875369727611542,
+      "learning_rate": 1.313499075630899e-05,
+      "loss": 0.0344,
+      "num_input_tokens_seen": 29195920,
+      "step": 710
+    },
+    {
+      "epoch": 12.935779816513762,
+      "grad_norm": 0.3268345892429352,
+      "learning_rate": 1.3071030993509788e-05,
+      "loss": 0.0357,
+      "num_input_tokens_seen": 29230376,
+      "step": 711
+    },
+    {
+      "epoch": 12.954128440366972,
+      "grad_norm": 1.2907904386520386,
+      "learning_rate": 1.3007172168743854e-05,
+      "loss": 0.058,
+      "num_input_tokens_seen": 29266464,
+      "step": 712
+    },
+    {
+      "epoch": 12.972477064220184,
+      "grad_norm": 0.29891443252563477,
+      "learning_rate": 1.2943414822358285e-05,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 29308840,
+      "step": 713
+    },
+    {
+      "epoch": 12.990825688073395,
+      "grad_norm": 0.3546947240829468,
+      "learning_rate": 1.2879759493841575e-05,
+      "loss": 0.0387,
+      "num_input_tokens_seen": 29346656,
+      "step": 714
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.4480898678302765,
+      "learning_rate": 1.2816206721818944e-05,
+      "loss": 0.0294,
+      "num_input_tokens_seen": 29368144,
+      "step": 715
+    },
+    {
+      "epoch": 13.01834862385321,
+      "grad_norm": 0.27912405133247375,
+      "learning_rate": 1.2752757044047827e-05,
+      "loss": 0.029,
+      "num_input_tokens_seen": 29406696,
+      "step": 716
+    },
+    {
+      "epoch": 13.036697247706423,
+      "grad_norm": 0.2766615152359009,
+      "learning_rate": 1.2689410997413325e-05,
+      "loss": 0.0292,
+      "num_input_tokens_seen": 29447448,
+      "step": 717
+    },
+    {
+      "epoch": 13.055045871559633,
+      "grad_norm": 0.29729965329170227,
+      "learning_rate": 1.262616911792365e-05,
+      "loss": 0.031,
+      "num_input_tokens_seen": 29493904,
+      "step": 718
+    },
+    {
+      "epoch": 13.073394495412844,
+      "grad_norm": 0.6886661052703857,
+      "learning_rate": 1.2563031940705594e-05,
+      "loss": 0.033,
+      "num_input_tokens_seen": 29533272,
+      "step": 719
+    },
+    {
+      "epoch": 13.091743119266056,
+      "grad_norm": 0.30814382433891296,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 0.0336,
+      "num_input_tokens_seen": 29565728,
+      "step": 720
+    },
+    {
+      "epoch": 13.110091743119266,
+      "grad_norm": 0.2744746804237366,
+      "learning_rate": 1.243707382915725e-05,
+      "loss": 0.0237,
+      "num_input_tokens_seen": 29604480,
+      "step": 721
+    },
+    {
+      "epoch": 13.128440366972477,
+      "grad_norm": 0.29700982570648193,
+      "learning_rate": 1.2374253960632757e-05,
+      "loss": 0.0282,
+      "num_input_tokens_seen": 29639856,
+      "step": 722
+    },
+    {
+      "epoch": 13.146788990825687,
+      "grad_norm": 0.33870425820350647,
+      "learning_rate": 1.2311540925982403e-05,
+      "loss": 0.0324,
+      "num_input_tokens_seen": 29677448,
+      "step": 723
+    },
+    {
+      "epoch": 13.1651376146789,
+      "grad_norm": 0.2735939025878906,
+      "learning_rate": 1.2248935255858117e-05,
+      "loss": 0.0269,
+      "num_input_tokens_seen": 29723888,
+      "step": 724
+    },
+    {
+      "epoch": 13.18348623853211,
+      "grad_norm": 0.2972582280635834,
+      "learning_rate": 1.2186437480003372e-05,
+      "loss": 0.0252,
+      "num_input_tokens_seen": 29767992,
+      "step": 725
+    },
+    {
+      "epoch": 13.20183486238532,
+      "grad_norm": 0.33630624413490295,
+      "learning_rate": 1.2124048127248644e-05,
+      "loss": 0.0386,
+      "num_input_tokens_seen": 29814304,
+      "step": 726
+    },
+    {
+      "epoch": 13.220183486238533,
+      "grad_norm": 0.351375013589859,
+      "learning_rate": 1.2061767725507006e-05,
+      "loss": 0.0336,
+      "num_input_tokens_seen": 29852040,
+      "step": 727
+    },
+    {
+      "epoch": 13.238532110091743,
+      "grad_norm": 0.3030643165111542,
+      "learning_rate": 1.1999596801769616e-05,
+      "loss": 0.0262,
+      "num_input_tokens_seen": 29889096,
+      "step": 728
+    },
+    {
+      "epoch": 13.256880733944953,
+      "grad_norm": 0.2586018443107605,
+      "learning_rate": 1.1937535882101281e-05,
+      "loss": 0.0251,
+      "num_input_tokens_seen": 29927888,
+      "step": 729
+    },
+    {
+      "epoch": 13.275229357798166,
+      "grad_norm": 0.2849186956882477,
+      "learning_rate": 1.1875585491636e-05,
+      "loss": 0.0315,
+      "num_input_tokens_seen": 29967064,
+      "step": 730
+    },
+    {
+      "epoch": 13.293577981651376,
+      "grad_norm": 0.26484936475753784,
+      "learning_rate": 1.1813746154572514e-05,
+      "loss": 0.0259,
+      "num_input_tokens_seen": 30016736,
+      "step": 731
+    },
+    {
+      "epoch": 13.311926605504587,
+      "grad_norm": 0.3491249084472656,
+      "learning_rate": 1.175201839416988e-05,
+      "loss": 0.0274,
+      "num_input_tokens_seen": 30056888,
+      "step": 732
+    },
+    {
+      "epoch": 13.330275229357799,
+      "grad_norm": 0.3179028630256653,
+      "learning_rate": 1.1690402732743042e-05,
+      "loss": 0.0272,
+      "num_input_tokens_seen": 30088720,
+      "step": 733
+    },
+    {
+      "epoch": 13.34862385321101,
+      "grad_norm": 0.9645937085151672,
+      "learning_rate": 1.1628899691658399e-05,
+      "loss": 0.0443,
+      "num_input_tokens_seen": 30131848,
+      "step": 734
+    },
+    {
+      "epoch": 13.36697247706422,
+      "grad_norm": 0.2913883626461029,
+      "learning_rate": 1.1567509791329401e-05,
+      "loss": 0.032,
+      "num_input_tokens_seen": 30186336,
+      "step": 735
+    },
+    {
+      "epoch": 13.385321100917432,
+      "grad_norm": 0.269525408744812,
+      "learning_rate": 1.1506233551212186e-05,
+      "loss": 0.0234,
+      "num_input_tokens_seen": 30242096,
+      "step": 736
+    },
+    {
+      "epoch": 13.403669724770642,
+      "grad_norm": 0.30820366740226746,
+      "learning_rate": 1.1445071489801073e-05,
+      "loss": 0.0242,
+      "num_input_tokens_seen": 30273336,
+      "step": 737
+    },
+    {
+      "epoch": 13.422018348623853,
+      "grad_norm": 0.31679078936576843,
+      "learning_rate": 1.1384024124624324e-05,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 30300544,
+      "step": 738
+    },
+    {
+      "epoch": 13.440366972477065,
+      "grad_norm": 0.2609107196331024,
+      "learning_rate": 1.1323091972239635e-05,
+      "loss": 0.0223,
+      "num_input_tokens_seen": 30343952,
+      "step": 739
+    },
+    {
+      "epoch": 13.458715596330276,
+      "grad_norm": 0.31270915269851685,
+      "learning_rate": 1.126227554822985e-05,
+      "loss": 0.0237,
+      "num_input_tokens_seen": 30393960,
+      "step": 740
+    },
+    {
+      "epoch": 13.477064220183486,
+      "grad_norm": 0.31794360280036926,
+      "learning_rate": 1.1201575367198547e-05,
+      "loss": 0.0513,
+      "num_input_tokens_seen": 30441256,
+      "step": 741
+    },
+    {
+      "epoch": 13.495412844036696,
+      "grad_norm": 0.2859482169151306,
+      "learning_rate": 1.1140991942765713e-05,
+      "loss": 0.0276,
+      "num_input_tokens_seen": 30480016,
+      "step": 742
+    },
+    {
+      "epoch": 13.513761467889909,
+      "grad_norm": 0.666972815990448,
+      "learning_rate": 1.1080525787563393e-05,
+      "loss": 0.0544,
+      "num_input_tokens_seen": 30531440,
+      "step": 743
+    },
+    {
+      "epoch": 13.53211009174312,
+      "grad_norm": 0.3061029613018036,
+      "learning_rate": 1.1020177413231334e-05,
+      "loss": 0.0252,
+      "num_input_tokens_seen": 30572720,
+      "step": 744
+    },
+    {
+      "epoch": 13.55045871559633,
+      "grad_norm": 0.3531540036201477,
+      "learning_rate": 1.0959947330412682e-05,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 30610416,
+      "step": 745
+    },
+    {
+      "epoch": 13.568807339449542,
+      "grad_norm": 0.2804684638977051,
+      "learning_rate": 1.0899836048749645e-05,
+      "loss": 0.0232,
+      "num_input_tokens_seen": 30656448,
+      "step": 746
+    },
+    {
+      "epoch": 13.587155963302752,
+      "grad_norm": 0.32002997398376465,
+      "learning_rate": 1.0839844076879185e-05,
+      "loss": 0.0277,
+      "num_input_tokens_seen": 30685584,
+      "step": 747
+    },
+    {
+      "epoch": 13.605504587155963,
+      "grad_norm": 0.2907179296016693,
+      "learning_rate": 1.0779971922428711e-05,
+      "loss": 0.0289,
+      "num_input_tokens_seen": 30729472,
+      "step": 748
+    },
+    {
+      "epoch": 13.623853211009175,
+      "grad_norm": 0.45727285742759705,
+      "learning_rate": 1.0720220092011782e-05,
+      "loss": 0.0371,
+      "num_input_tokens_seen": 30765400,
+      "step": 749
+    },
+    {
+      "epoch": 13.642201834862385,
+      "grad_norm": 0.31665271520614624,
+      "learning_rate": 1.0660589091223855e-05,
+      "loss": 0.032,
+      "num_input_tokens_seen": 30812288,
+      "step": 750
+    },
+    {
+      "epoch": 13.660550458715596,
+      "grad_norm": 0.31468474864959717,
+      "learning_rate": 1.0601079424637917e-05,
+      "loss": 0.027,
+      "num_input_tokens_seen": 30855592,
+      "step": 751
+    },
+    {
+      "epoch": 13.678899082568808,
+      "grad_norm": 0.32163867354393005,
+      "learning_rate": 1.0541691595800337e-05,
+      "loss": 0.0337,
+      "num_input_tokens_seen": 30890072,
+      "step": 752
+    },
+    {
+      "epoch": 13.697247706422019,
+      "grad_norm": 0.26351386308670044,
+      "learning_rate": 1.0482426107226507e-05,
+      "loss": 0.0238,
+      "num_input_tokens_seen": 30927336,
+      "step": 753
+    },
+    {
+      "epoch": 13.715596330275229,
+      "grad_norm": 0.3523198664188385,
+      "learning_rate": 1.0423283460396633e-05,
+      "loss": 0.0294,
+      "num_input_tokens_seen": 30966920,
+      "step": 754
+    },
+    {
+      "epoch": 13.73394495412844,
+      "grad_norm": 0.262052059173584,
+      "learning_rate": 1.0364264155751488e-05,
+      "loss": 0.0281,
+      "num_input_tokens_seen": 31015544,
+      "step": 755
+    },
+    {
+      "epoch": 13.752293577981652,
+      "grad_norm": 0.3480219542980194,
+      "learning_rate": 1.0305368692688174e-05,
+      "loss": 0.0311,
+      "num_input_tokens_seen": 31052688,
+      "step": 756
+    },
+    {
+      "epoch": 13.770642201834862,
+      "grad_norm": 0.27989810705184937,
+      "learning_rate": 1.0246597569555894e-05,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 31092736,
+      "step": 757
+    },
+    {
+      "epoch": 13.788990825688073,
+      "grad_norm": 0.40765848755836487,
+      "learning_rate": 1.0187951283651736e-05,
+      "loss": 0.038,
+      "num_input_tokens_seen": 31123536,
+      "step": 758
+    },
+    {
+      "epoch": 13.807339449541285,
+      "grad_norm": 0.3290911316871643,
+      "learning_rate": 1.0129430331216471e-05,
+      "loss": 0.0321,
+      "num_input_tokens_seen": 31157144,
+      "step": 759
+    },
+    {
+      "epoch": 13.825688073394495,
+      "grad_norm": 0.2592953145503998,
+      "learning_rate": 1.0071035207430352e-05,
+      "loss": 0.0267,
+      "num_input_tokens_seen": 31222320,
+      "step": 760
+    },
+    {
+      "epoch": 13.844036697247706,
+      "grad_norm": 0.2886260449886322,
+      "learning_rate": 1.001276640640891e-05,
+      "loss": 0.0261,
+      "num_input_tokens_seen": 31277880,
+      "step": 761
+    },
+    {
+      "epoch": 13.862385321100918,
+      "grad_norm": 0.6212909817695618,
+      "learning_rate": 9.954624421198792e-06,
+      "loss": 0.0639,
+      "num_input_tokens_seen": 31338856,
+      "step": 762
+    },
+    {
+      "epoch": 13.880733944954128,
+      "grad_norm": 0.33612295985221863,
+      "learning_rate": 9.89660974377359e-06,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 31376304,
+      "step": 763
+    },
+    {
+      "epoch": 13.899082568807339,
+      "grad_norm": 0.3871086835861206,
+      "learning_rate": 9.838722865029673e-06,
+      "loss": 0.0371,
+      "num_input_tokens_seen": 31410672,
+      "step": 764
+    },
+    {
+      "epoch": 13.917431192660551,
+      "grad_norm": 0.3187819719314575,
+      "learning_rate": 9.780964274781984e-06,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 31444928,
+      "step": 765
+    },
+    {
+      "epoch": 13.935779816513762,
+      "grad_norm": 0.32425767183303833,
+      "learning_rate": 9.723334461760006e-06,
+      "loss": 0.0357,
+      "num_input_tokens_seen": 31482232,
+      "step": 766
+    },
+    {
+      "epoch": 13.954128440366972,
+      "grad_norm": 0.3090585470199585,
+      "learning_rate": 9.665833913603523e-06,
+      "loss": 0.0312,
+      "num_input_tokens_seen": 31513856,
+      "step": 767
+    },
+    {
+      "epoch": 13.972477064220184,
+      "grad_norm": 0.36073872447013855,
+      "learning_rate": 9.608463116858542e-06,
+      "loss": 0.0584,
+      "num_input_tokens_seen": 31563800,
+      "step": 768
+    },
+    {
+      "epoch": 13.990825688073395,
+      "grad_norm": 0.3506834805011749,
+      "learning_rate": 9.551222556973172e-06,
+      "loss": 0.0356,
+      "num_input_tokens_seen": 31606368,
+      "step": 769
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.5738533735275269,
+      "learning_rate": 9.494112718293501e-06,
+      "loss": 0.0511,
+      "num_input_tokens_seen": 31627232,
+      "step": 770
+    },
+    {
+      "epoch": 14.01834862385321,
+      "grad_norm": 0.24552291631698608,
+      "learning_rate": 9.437134084059515e-06,
+      "loss": 0.0237,
+      "num_input_tokens_seen": 31669296,
+      "step": 771
+    },
+    {
+      "epoch": 14.036697247706423,
+      "grad_norm": 0.25540077686309814,
+      "learning_rate": 9.380287136401e-06,
+      "loss": 0.026,
+      "num_input_tokens_seen": 31704400,
+      "step": 772
+    },
+    {
+      "epoch": 14.055045871559633,
+      "grad_norm": 0.293837308883667,
+      "learning_rate": 9.323572356333454e-06,
+      "loss": 0.0252,
+      "num_input_tokens_seen": 31739048,
+      "step": 773
+    },
+    {
+      "epoch": 14.073394495412844,
+      "grad_norm": 0.2588525116443634,
+      "learning_rate": 9.266990223754069e-06,
+      "loss": 0.0175,
+      "num_input_tokens_seen": 31776056,
+      "step": 774
+    },
+    {
+      "epoch": 14.091743119266056,
+      "grad_norm": 0.28983962535858154,
+      "learning_rate": 9.210541217437565e-06,
+      "loss": 0.0269,
+      "num_input_tokens_seen": 31814504,
+      "step": 775
+    },
+    {
+      "epoch": 14.110091743119266,
+      "grad_norm": 0.32418689131736755,
+      "learning_rate": 9.154225815032242e-06,
+      "loss": 0.0283,
+      "num_input_tokens_seen": 31846928,
+      "step": 776
+    },
+    {
+      "epoch": 14.128440366972477,
+      "grad_norm": 0.26463258266448975,
+      "learning_rate": 9.098044493055899e-06,
+      "loss": 0.0256,
+      "num_input_tokens_seen": 31893152,
+      "step": 777
+    },
+    {
+      "epoch": 14.146788990825687,
+      "grad_norm": 0.24525673687458038,
+      "learning_rate": 9.0419977268918e-06,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 31937456,
+      "step": 778
+    },
+    {
+      "epoch": 14.1651376146789,
+      "grad_norm": 0.300974577665329,
+      "learning_rate": 8.98608599078462e-06,
+      "loss": 0.0301,
+      "num_input_tokens_seen": 31969928,
+      "step": 779
+    },
+    {
+      "epoch": 14.18348623853211,
+      "grad_norm": 0.31297603249549866,
+      "learning_rate": 8.930309757836517e-06,
+      "loss": 0.0256,
+      "num_input_tokens_seen": 32032920,
+      "step": 780
+    },
+    {
+      "epoch": 14.20183486238532,
+      "grad_norm": 0.2529805898666382,
+      "learning_rate": 8.874669500003049e-06,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 32087928,
+      "step": 781
+    },
+    {
+      "epoch": 14.220183486238533,
+      "grad_norm": 0.2599085867404938,
+      "learning_rate": 8.819165688089193e-06,
+      "loss": 0.026,
+      "num_input_tokens_seen": 32133632,
+      "step": 782
+    },
+    {
+      "epoch": 14.238532110091743,
+      "grad_norm": 0.3724299967288971,
+      "learning_rate": 8.763798791745411e-06,
+      "loss": 0.0307,
+      "num_input_tokens_seen": 32178024,
+      "step": 783
+    },
+    {
+      "epoch": 14.256880733944953,
+      "grad_norm": 0.29776036739349365,
+      "learning_rate": 8.708569279463622e-06,
+      "loss": 0.028,
+      "num_input_tokens_seen": 32229216,
+      "step": 784
+    },
+    {
+      "epoch": 14.275229357798166,
+      "grad_norm": 0.326056569814682,
+      "learning_rate": 8.65347761857326e-06,
+      "loss": 0.047,
+      "num_input_tokens_seen": 32279544,
+      "step": 785
+    },
+    {
+      "epoch": 14.293577981651376,
+      "grad_norm": 0.24950319528579712,
+      "learning_rate": 8.598524275237322e-06,
+      "loss": 0.0221,
+      "num_input_tokens_seen": 32314736,
+      "step": 786
+    },
+    {
+      "epoch": 14.311926605504587,
+      "grad_norm": 0.2621147632598877,
+      "learning_rate": 8.543709714448403e-06,
+      "loss": 0.0211,
+      "num_input_tokens_seen": 32357232,
+      "step": 787
+    },
+    {
+      "epoch": 14.330275229357799,
+      "grad_norm": 0.3251303732395172,
+      "learning_rate": 8.489034400024812e-06,
+      "loss": 0.0304,
+      "num_input_tokens_seen": 32392992,
+      "step": 788
+    },
+    {
+      "epoch": 14.34862385321101,
+      "grad_norm": 0.26626235246658325,
+      "learning_rate": 8.434498794606568e-06,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 32434840,
+      "step": 789
+    },
+    {
+      "epoch": 14.36697247706422,
+      "grad_norm": 0.24833904206752777,
+      "learning_rate": 8.380103359651553e-06,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 32480592,
+      "step": 790
+    },
+    {
+      "epoch": 14.385321100917432,
+      "grad_norm": 0.6969995498657227,
+      "learning_rate": 8.325848555431595e-06,
+      "loss": 0.0258,
+      "num_input_tokens_seen": 32514440,
+      "step": 791
+    },
+    {
+      "epoch": 14.403669724770642,
+      "grad_norm": 0.30866381525993347,
+      "learning_rate": 8.271734841028553e-06,
+      "loss": 0.045,
+      "num_input_tokens_seen": 32557176,
+      "step": 792
+    },
+    {
+      "epoch": 14.422018348623853,
+      "grad_norm": 0.33566537499427795,
+      "learning_rate": 8.217762674330413e-06,
+      "loss": 0.0304,
+      "num_input_tokens_seen": 32594416,
+      "step": 793
+    },
+    {
+      "epoch": 14.440366972477065,
+      "grad_norm": 0.23681727051734924,
+      "learning_rate": 8.163932512027492e-06,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 32638344,
+      "step": 794
+    },
+    {
+      "epoch": 14.458715596330276,
+      "grad_norm": 0.3150688111782074,
+      "learning_rate": 8.110244809608495e-06,
+      "loss": 0.0336,
+      "num_input_tokens_seen": 32674416,
+      "step": 795
+    },
+    {
+      "epoch": 14.477064220183486,
+      "grad_norm": 0.2548944354057312,
+      "learning_rate": 8.056700021356694e-06,
+      "loss": 0.0309,
+      "num_input_tokens_seen": 32725808,
+      "step": 796
+    },
+    {
+      "epoch": 14.495412844036696,
+      "grad_norm": 0.3117126524448395,
+      "learning_rate": 8.003298600346085e-06,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 32761224,
+      "step": 797
+    },
+    {
+      "epoch": 14.513761467889909,
+      "grad_norm": 0.2600093483924866,
+      "learning_rate": 7.950040998437542e-06,
+      "loss": 0.0235,
+      "num_input_tokens_seen": 32812208,
+      "step": 798
+    },
+    {
+      "epoch": 14.53211009174312,
+      "grad_norm": 0.24391524493694305,
+      "learning_rate": 7.896927666275006e-06,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 32858672,
+      "step": 799
+    },
+    {
+      "epoch": 14.55045871559633,
+      "grad_norm": 0.3656960129737854,
+      "learning_rate": 7.843959053281663e-06,
+      "loss": 0.0397,
+      "num_input_tokens_seen": 32905792,
+      "step": 800
+    },
+    {
+      "epoch": 14.568807339449542,
+      "grad_norm": 0.249402716755867,
+      "learning_rate": 7.791135607656147e-06,
+      "loss": 0.0216,
+      "num_input_tokens_seen": 32946456,
+      "step": 801
+    },
+    {
+      "epoch": 14.587155963302752,
+      "grad_norm": 0.2628585696220398,
+      "learning_rate": 7.738457776368766e-06,
+      "loss": 0.0201,
+      "num_input_tokens_seen": 32985008,
+      "step": 802
+    },
+    {
+      "epoch": 14.605504587155963,
+      "grad_norm": 0.26252928376197815,
+      "learning_rate": 7.685926005157651e-06,
+      "loss": 0.0229,
+      "num_input_tokens_seen": 33029600,
+      "step": 803
+    },
+    {
+      "epoch": 14.623853211009175,
+      "grad_norm": 0.33849260210990906,
+      "learning_rate": 7.633540738525066e-06,
+      "loss": 0.0268,
+      "num_input_tokens_seen": 33064056,
+      "step": 804
+    },
+    {
+      "epoch": 14.642201834862385,
+      "grad_norm": 0.2663592994213104,
+      "learning_rate": 7.581302419733632e-06,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 33099880,
+      "step": 805
+    },
+    {
+      "epoch": 14.660550458715596,
+      "grad_norm": 0.2511848211288452,
+      "learning_rate": 7.529211490802498e-06,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 33134376,
+      "step": 806
+    },
+    {
+      "epoch": 14.678899082568808,
+      "grad_norm": 0.2512349784374237,
+      "learning_rate": 7.477268392503728e-06,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 33172056,
+      "step": 807
+    },
+    {
+      "epoch": 14.697247706422019,
+      "grad_norm": 0.2771822512149811,
+      "learning_rate": 7.4254735643584564e-06,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 33224448,
+      "step": 808
+    },
+    {
+      "epoch": 14.715596330275229,
+      "grad_norm": 0.24601776897907257,
+      "learning_rate": 7.3738274446332415e-06,
+      "loss": 0.0228,
+      "num_input_tokens_seen": 33264800,
+      "step": 809
+    },
+    {
+      "epoch": 14.73394495412844,
+      "grad_norm": 0.33897870779037476,
+      "learning_rate": 7.3223304703363135e-06,
+      "loss": 0.025,
+      "num_input_tokens_seen": 33300896,
+      "step": 810
+    },
+    {
+      "epoch": 14.752293577981652,
+      "grad_norm": 0.2866343855857849,
+      "learning_rate": 7.270983077213911e-06,
+      "loss": 0.0243,
+      "num_input_tokens_seen": 33335304,
+      "step": 811
+    },
+    {
+      "epoch": 14.770642201834862,
+      "grad_norm": 0.22896356880664825,
+      "learning_rate": 7.219785699746573e-06,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 33382912,
+      "step": 812
+    },
+    {
+      "epoch": 14.788990825688073,
+      "grad_norm": 0.2809698283672333,
+      "learning_rate": 7.168738771145464e-06,
+      "loss": 0.0299,
+      "num_input_tokens_seen": 33422688,
+      "step": 813
+    },
+    {
+      "epoch": 14.807339449541285,
+      "grad_norm": 0.32832175493240356,
+      "learning_rate": 7.117842723348717e-06,
+      "loss": 0.0301,
+      "num_input_tokens_seen": 33456520,
+      "step": 814
+    },
+    {
+      "epoch": 14.825688073394495,
+      "grad_norm": 0.3265829384326935,
+      "learning_rate": 7.067097987017762e-06,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 33501112,
+      "step": 815
+    },
+    {
+      "epoch": 14.844036697247706,
+      "grad_norm": 0.3356791138648987,
+      "learning_rate": 7.016504991533726e-06,
+      "loss": 0.0314,
+      "num_input_tokens_seen": 33535160,
+      "step": 816
+    },
+    {
+      "epoch": 14.862385321100918,
+      "grad_norm": 0.28199252486228943,
+      "learning_rate": 6.9660641649937155e-06,
+      "loss": 0.0289,
+      "num_input_tokens_seen": 33583896,
+      "step": 817
+    },
+    {
+      "epoch": 14.880733944954128,
+      "grad_norm": 0.28438156843185425,
+      "learning_rate": 6.9157759342072995e-06,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 33616680,
+      "step": 818
+    },
+    {
+      "epoch": 14.899082568807339,
+      "grad_norm": 0.35118597745895386,
+      "learning_rate": 6.865640724692815e-06,
+      "loss": 0.029,
+      "num_input_tokens_seen": 33652496,
+      "step": 819
+    },
+    {
+      "epoch": 14.917431192660551,
+      "grad_norm": 0.27889585494995117,
+      "learning_rate": 6.815658960673782e-06,
+      "loss": 0.0379,
+      "num_input_tokens_seen": 33695760,
+      "step": 820
+    },
+    {
+      "epoch": 14.935779816513762,
+      "grad_norm": 0.26556557416915894,
+      "learning_rate": 6.765831065075367e-06,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 33734472,
+      "step": 821
+    },
+    {
+      "epoch": 14.954128440366972,
+      "grad_norm": 0.22440935671329498,
+      "learning_rate": 6.716157459520739e-06,
+      "loss": 0.0243,
+      "num_input_tokens_seen": 33792456,
+      "step": 822
+    },
+    {
+      "epoch": 14.972477064220184,
+      "grad_norm": 0.24052810668945312,
+      "learning_rate": 6.666638564327532e-06,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 33837616,
+      "step": 823
+    },
+    {
+      "epoch": 14.990825688073395,
+      "grad_norm": 0.3531548082828522,
+      "learning_rate": 6.617274798504286e-06,
+      "loss": 0.0316,
+      "num_input_tokens_seen": 33871912,
+      "step": 824
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.4734267592430115,
+      "learning_rate": 6.568066579746901e-06,
+      "loss": 0.038,
+      "num_input_tokens_seen": 33886320,
+      "step": 825
+    },
+    {
+      "epoch": 15.01834862385321,
+      "grad_norm": 0.22462400794029236,
+      "learning_rate": 6.519014324435102e-06,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 33929160,
+      "step": 826
+    },
+    {
+      "epoch": 15.036697247706423,
+      "grad_norm": 0.29892125725746155,
+      "learning_rate": 6.470118447628912e-06,
+      "loss": 0.0303,
+      "num_input_tokens_seen": 33976056,
+      "step": 827
+    },
+    {
+      "epoch": 15.055045871559633,
+      "grad_norm": 0.2654046416282654,
+      "learning_rate": 6.421379363065142e-06,
+      "loss": 0.0394,
+      "num_input_tokens_seen": 34016992,
+      "step": 828
+    },
+    {
+      "epoch": 15.073394495412844,
+      "grad_norm": 0.24058304727077484,
+      "learning_rate": 6.3727974831539115e-06,
+      "loss": 0.0238,
+      "num_input_tokens_seen": 34054944,
+      "step": 829
+    },
+    {
+      "epoch": 15.091743119266056,
+      "grad_norm": 0.2786811888217926,
+      "learning_rate": 6.324373218975105e-06,
+      "loss": 0.0382,
+      "num_input_tokens_seen": 34100400,
+      "step": 830
+    },
+    {
+      "epoch": 15.110091743119266,
+      "grad_norm": 0.24469152092933655,
+      "learning_rate": 6.2761069802749455e-06,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 34154080,
+      "step": 831
+    },
+    {
+      "epoch": 15.128440366972477,
+      "grad_norm": 0.21530218422412872,
+      "learning_rate": 6.22799917546252e-06,
+      "loss": 0.0192,
+      "num_input_tokens_seen": 34211416,
+      "step": 832
+    },
+    {
+      "epoch": 15.146788990825687,
+      "grad_norm": 0.25735238194465637,
+      "learning_rate": 6.180050211606303e-06,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 34250208,
+      "step": 833
+    },
+    {
+      "epoch": 15.1651376146789,
+      "grad_norm": 0.2362961620092392,
+      "learning_rate": 6.1322604944307e-06,
+      "loss": 0.0188,
+      "num_input_tokens_seen": 34305792,
+      "step": 834
+    },
+    {
+      "epoch": 15.18348623853211,
+      "grad_norm": 0.23253802955150604,
+      "learning_rate": 6.08463042831268e-06,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 34345272,
+      "step": 835
+    },
+    {
+      "epoch": 15.20183486238532,
+      "grad_norm": 0.2991214394569397,
+      "learning_rate": 6.037160416278278e-06,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 34381752,
+      "step": 836
+    },
+    {
+      "epoch": 15.220183486238533,
+      "grad_norm": 0.257041335105896,
+      "learning_rate": 5.989850859999227e-06,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 34423552,
+      "step": 837
+    },
+    {
+      "epoch": 15.238532110091743,
+      "grad_norm": 0.22987478971481323,
+      "learning_rate": 5.942702159789554e-06,
+      "loss": 0.0204,
+      "num_input_tokens_seen": 34463296,
+      "step": 838
+    },
+    {
+      "epoch": 15.256880733944953,
+      "grad_norm": 0.3145504295825958,
+      "learning_rate": 5.895714714602171e-06,
+      "loss": 0.0262,
+      "num_input_tokens_seen": 34494592,
+      "step": 839
+    },
+    {
+      "epoch": 15.275229357798166,
+      "grad_norm": 0.2484779804944992,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 0.0232,
+      "num_input_tokens_seen": 34536560,
+      "step": 840
+    },
+    {
+      "epoch": 15.293577981651376,
+      "grad_norm": 0.28855112195014954,
+      "learning_rate": 5.802225178280288e-06,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 34565424,
+      "step": 841
+    },
+    {
+      "epoch": 15.311926605504587,
+      "grad_norm": 0.21636433899402618,
+      "learning_rate": 5.755723878215802e-06,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 34615248,
+      "step": 842
+    },
+    {
+      "epoch": 15.330275229357799,
+      "grad_norm": 0.30022677779197693,
+      "learning_rate": 5.709385415307006e-06,
+      "loss": 0.0237,
+      "num_input_tokens_seen": 34643648,
+      "step": 843
+    },
+    {
+      "epoch": 15.34862385321101,
+      "grad_norm": 0.23669451475143433,
+      "learning_rate": 5.663210181650905e-06,
+      "loss": 0.0189,
+      "num_input_tokens_seen": 34678888,
+      "step": 844
+    },
+    {
+      "epoch": 15.36697247706422,
+      "grad_norm": 0.3009994924068451,
+      "learning_rate": 5.617198567963352e-06,
+      "loss": 0.0228,
+      "num_input_tokens_seen": 34717536,
+      "step": 845
+    },
+    {
+      "epoch": 15.385321100917432,
+      "grad_norm": 0.2476666420698166,
+      "learning_rate": 5.571350963575728e-06,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 34755840,
+      "step": 846
+    },
+    {
+      "epoch": 15.403669724770642,
+      "grad_norm": 0.2747964859008789,
+      "learning_rate": 5.525667756431616e-06,
+      "loss": 0.0277,
+      "num_input_tokens_seen": 34788648,
+      "step": 847
+    },
+    {
+      "epoch": 15.422018348623853,
+      "grad_norm": 0.3019561469554901,
+      "learning_rate": 5.48014933308352e-06,
+      "loss": 0.0261,
+      "num_input_tokens_seen": 34818368,
+      "step": 848
+    },
+    {
+      "epoch": 15.440366972477065,
+      "grad_norm": 0.22335736453533173,
+      "learning_rate": 5.434796078689652e-06,
+      "loss": 0.0243,
+      "num_input_tokens_seen": 34877832,
+      "step": 849
+    },
+    {
+      "epoch": 15.458715596330276,
+      "grad_norm": 0.34833160042762756,
+      "learning_rate": 5.389608377010608e-06,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 34929560,
+      "step": 850
+    },
+    {
+      "epoch": 15.477064220183486,
+      "grad_norm": 0.2435968518257141,
+      "learning_rate": 5.344586610406146e-06,
+      "loss": 0.0198,
+      "num_input_tokens_seen": 34969976,
+      "step": 851
+    },
+    {
+      "epoch": 15.495412844036696,
+      "grad_norm": 0.22303880751132965,
+      "learning_rate": 5.299731159831953e-06,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 35008288,
+      "step": 852
+    },
+    {
+      "epoch": 15.513761467889909,
+      "grad_norm": 0.30280816555023193,
+      "learning_rate": 5.2550424048364185e-06,
+      "loss": 0.0314,
+      "num_input_tokens_seen": 35045016,
+      "step": 853
+    },
+    {
+      "epoch": 15.53211009174312,
+      "grad_norm": 0.2565707266330719,
+      "learning_rate": 5.210520723557419e-06,
+      "loss": 0.0192,
+      "num_input_tokens_seen": 35085312,
+      "step": 854
+    },
+    {
+      "epoch": 15.55045871559633,
+      "grad_norm": 0.24315547943115234,
+      "learning_rate": 5.166166492719124e-06,
+      "loss": 0.0243,
+      "num_input_tokens_seen": 35130040,
+      "step": 855
+    },
+    {
+      "epoch": 15.568807339449542,
+      "grad_norm": 0.30669093132019043,
+      "learning_rate": 5.121980087628803e-06,
+      "loss": 0.026,
+      "num_input_tokens_seen": 35169456,
+      "step": 856
+    },
+    {
+      "epoch": 15.587155963302752,
+      "grad_norm": 0.207157701253891,
+      "learning_rate": 5.077961882173676e-06,
+      "loss": 0.0177,
+      "num_input_tokens_seen": 35216632,
+      "step": 857
+    },
+    {
+      "epoch": 15.605504587155963,
+      "grad_norm": 0.2158558964729309,
+      "learning_rate": 5.034112248817685e-06,
+      "loss": 0.0175,
+      "num_input_tokens_seen": 35263048,
+      "step": 858
+    },
+    {
+      "epoch": 15.623853211009175,
+      "grad_norm": 0.2432718127965927,
+      "learning_rate": 4.990431558598408e-06,
+      "loss": 0.0182,
+      "num_input_tokens_seen": 35306376,
+      "step": 859
+    },
+    {
+      "epoch": 15.642201834862385,
+      "grad_norm": 0.2845425009727478,
+      "learning_rate": 4.946920181123904e-06,
+      "loss": 0.0236,
+      "num_input_tokens_seen": 35341168,
+      "step": 860
+    },
+    {
+      "epoch": 15.660550458715596,
+      "grad_norm": 0.23385128378868103,
+      "learning_rate": 4.903578484569568e-06,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 35376920,
+      "step": 861
+    },
+    {
+      "epoch": 15.678899082568808,
+      "grad_norm": 0.2769944965839386,
+      "learning_rate": 4.860406835675016e-06,
+      "loss": 0.0257,
+      "num_input_tokens_seen": 35430304,
+      "step": 862
+    },
+    {
+      "epoch": 15.697247706422019,
+      "grad_norm": 0.24330423772335052,
+      "learning_rate": 4.817405599741004e-06,
+      "loss": 0.0218,
+      "num_input_tokens_seen": 35466728,
+      "step": 863
+    },
+    {
+      "epoch": 15.715596330275229,
+      "grad_norm": 0.25154146552085876,
+      "learning_rate": 4.7745751406263165e-06,
+      "loss": 0.0231,
+      "num_input_tokens_seen": 35503792,
+      "step": 864
+    },
+    {
+      "epoch": 15.73394495412844,
+      "grad_norm": 0.31640398502349854,
+      "learning_rate": 4.731915820744695e-06,
+      "loss": 0.0372,
+      "num_input_tokens_seen": 35545312,
+      "step": 865
+    },
+    {
+      "epoch": 15.752293577981652,
+      "grad_norm": 0.3148432672023773,
+      "learning_rate": 4.689428001061774e-06,
+      "loss": 0.0281,
+      "num_input_tokens_seen": 35576480,
+      "step": 866
+    },
+    {
+      "epoch": 15.770642201834862,
+      "grad_norm": 0.29388168454170227,
+      "learning_rate": 4.647112041092022e-06,
+      "loss": 0.0214,
+      "num_input_tokens_seen": 35613240,
+      "step": 867
+    },
+    {
+      "epoch": 15.788990825688073,
+      "grad_norm": 0.24217398464679718,
+      "learning_rate": 4.604968298895703e-06,
+      "loss": 0.0296,
+      "num_input_tokens_seen": 35671328,
+      "step": 868
+    },
+    {
+      "epoch": 15.807339449541285,
+      "grad_norm": 0.3247559666633606,
+      "learning_rate": 4.562997131075841e-06,
+      "loss": 0.0256,
+      "num_input_tokens_seen": 35701032,
+      "step": 869
+    },
+    {
+      "epoch": 15.825688073394495,
+      "grad_norm": 0.2986390292644501,
+      "learning_rate": 4.521198892775203e-06,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 35759112,
+      "step": 870
+    },
+    {
+      "epoch": 15.844036697247706,
+      "grad_norm": 0.2897619605064392,
+      "learning_rate": 4.479573937673326e-06,
+      "loss": 0.0222,
+      "num_input_tokens_seen": 35794232,
+      "step": 871
+    },
+    {
+      "epoch": 15.862385321100918,
+      "grad_norm": 0.2992808222770691,
+      "learning_rate": 4.438122617983443e-06,
+      "loss": 0.0266,
+      "num_input_tokens_seen": 35825360,
+      "step": 872
+    },
+    {
+      "epoch": 15.880733944954128,
+      "grad_norm": 0.27834197878837585,
+      "learning_rate": 4.396845284449608e-06,
+      "loss": 0.0238,
+      "num_input_tokens_seen": 35860784,
+      "step": 873
+    },
+    {
+      "epoch": 15.899082568807339,
+      "grad_norm": 0.2583276927471161,
+      "learning_rate": 4.355742286343648e-06,
+      "loss": 0.0243,
+      "num_input_tokens_seen": 35915640,
+      "step": 874
+    },
+    {
+      "epoch": 15.917431192660551,
+      "grad_norm": 0.2062636911869049,
+      "learning_rate": 4.314813971462237e-06,
+      "loss": 0.019,
+      "num_input_tokens_seen": 35970056,
+      "step": 875
+    },
+    {
+      "epoch": 15.935779816513762,
+      "grad_norm": 0.2623538076877594,
+      "learning_rate": 4.274060686123959e-06,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 36004072,
+      "step": 876
+    },
+    {
+      "epoch": 15.954128440366972,
+      "grad_norm": 0.26135191321372986,
+      "learning_rate": 4.233482775166364e-06,
+      "loss": 0.0199,
+      "num_input_tokens_seen": 36056968,
+      "step": 877
+    },
+    {
+      "epoch": 15.972477064220184,
+      "grad_norm": 0.3041679263114929,
+      "learning_rate": 4.19308058194306e-06,
+      "loss": 0.0256,
+      "num_input_tokens_seen": 36091736,
+      "step": 878
+    },
+    {
+      "epoch": 15.990825688073395,
+      "grad_norm": 0.296235054731369,
+      "learning_rate": 4.152854448320797e-06,
+      "loss": 0.0245,
+      "num_input_tokens_seen": 36129176,
+      "step": 879
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.4424769878387451,
+      "learning_rate": 4.112804714676594e-06,
+      "loss": 0.028,
+      "num_input_tokens_seen": 36145408,
+      "step": 880
+    },
+    {
+      "epoch": 16.01834862385321,
+      "grad_norm": 0.22620835900306702,
+      "learning_rate": 4.072931719894829e-06,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 36186960,
+      "step": 881
+    },
+    {
+      "epoch": 16.03669724770642,
+      "grad_norm": 0.20474722981452942,
+      "learning_rate": 4.0332358013644016e-06,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 36243640,
+      "step": 882
+    },
+    {
+      "epoch": 16.05504587155963,
+      "grad_norm": 0.22973741590976715,
+      "learning_rate": 3.993717294975863e-06,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 36285344,
+      "step": 883
+    },
+    {
+      "epoch": 16.073394495412845,
+      "grad_norm": 0.2423926442861557,
+      "learning_rate": 3.954376535118562e-06,
+      "loss": 0.0264,
+      "num_input_tokens_seen": 36316512,
+      "step": 884
+    },
+    {
+      "epoch": 16.091743119266056,
+      "grad_norm": 0.23595216870307922,
+      "learning_rate": 3.9152138546778625e-06,
+      "loss": 0.0182,
+      "num_input_tokens_seen": 36351392,
+      "step": 885
+    },
+    {
+      "epoch": 16.110091743119266,
+      "grad_norm": 0.25316476821899414,
+      "learning_rate": 3.876229585032245e-06,
+      "loss": 0.0198,
+      "num_input_tokens_seen": 36385144,
+      "step": 886
+    },
+    {
+      "epoch": 16.128440366972477,
+      "grad_norm": 0.25754794478416443,
+      "learning_rate": 3.837424056050598e-06,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 36423008,
+      "step": 887
+    },
+    {
+      "epoch": 16.146788990825687,
+      "grad_norm": 0.29055434465408325,
+      "learning_rate": 3.798797596089351e-06,
+      "loss": 0.0226,
+      "num_input_tokens_seen": 36454768,
+      "step": 888
+    },
+    {
+      "epoch": 16.165137614678898,
+      "grad_norm": 0.22524721920490265,
+      "learning_rate": 3.760350531989737e-06,
+      "loss": 0.0208,
+      "num_input_tokens_seen": 36503200,
+      "step": 889
+    },
+    {
+      "epoch": 16.18348623853211,
+      "grad_norm": 0.22587193548679352,
+      "learning_rate": 3.7220831890750067e-06,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 36544752,
+      "step": 890
+    },
+    {
+      "epoch": 16.201834862385322,
+      "grad_norm": 0.2937374413013458,
+      "learning_rate": 3.6839958911476957e-06,
+      "loss": 0.024,
+      "num_input_tokens_seen": 36589352,
+      "step": 891
+    },
+    {
+      "epoch": 16.220183486238533,
+      "grad_norm": 0.2755613625049591,
+      "learning_rate": 3.6460889604868626e-06,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 36619488,
+      "step": 892
+    },
+    {
+      "epoch": 16.238532110091743,
+      "grad_norm": 0.22830481827259064,
+      "learning_rate": 3.608362717845376e-06,
+      "loss": 0.023,
+      "num_input_tokens_seen": 36659792,
+      "step": 893
+    },
+    {
+      "epoch": 16.256880733944953,
+      "grad_norm": 0.2376595437526703,
+      "learning_rate": 3.5708174824471947e-06,
+      "loss": 0.0185,
+      "num_input_tokens_seen": 36700648,
+      "step": 894
+    },
+    {
+      "epoch": 16.275229357798164,
+      "grad_norm": 0.25321322679519653,
+      "learning_rate": 3.5334535719846766e-06,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 36755856,
+      "step": 895
+    },
+    {
+      "epoch": 16.293577981651374,
+      "grad_norm": 0.25597620010375977,
+      "learning_rate": 3.4962713026158694e-06,
+      "loss": 0.0204,
+      "num_input_tokens_seen": 36788392,
+      "step": 896
+    },
+    {
+      "epoch": 16.31192660550459,
+      "grad_norm": 0.25083833932876587,
+      "learning_rate": 3.4592709889618545e-06,
+      "loss": 0.0184,
+      "num_input_tokens_seen": 36821312,
+      "step": 897
+    },
+    {
+      "epoch": 16.3302752293578,
+      "grad_norm": 0.2870706617832184,
+      "learning_rate": 3.4224529441040904e-06,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 36859304,
+      "step": 898
+    },
+    {
+      "epoch": 16.34862385321101,
+      "grad_norm": 0.21498925983905792,
+      "learning_rate": 3.3858174795817404e-06,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 36905288,
+      "step": 899
+    },
+    {
+      "epoch": 16.36697247706422,
+      "grad_norm": 0.23456843197345734,
+      "learning_rate": 3.3493649053890326e-06,
+      "loss": 0.0207,
+      "num_input_tokens_seen": 36951576,
+      "step": 900
+    },
+    {
+      "epoch": 16.38532110091743,
+      "grad_norm": 0.2607397139072418,
+      "learning_rate": 3.313095529972687e-06,
+      "loss": 0.02,
+      "num_input_tokens_seen": 36985192,
+      "step": 901
+    },
+    {
+      "epoch": 16.40366972477064,
+      "grad_norm": 0.2284688800573349,
+      "learning_rate": 3.2770096602292465e-06,
+      "loss": 0.0232,
+      "num_input_tokens_seen": 37031592,
+      "step": 902
+    },
+    {
+      "epoch": 16.422018348623855,
+      "grad_norm": 0.2386486828327179,
+      "learning_rate": 3.2411076015025075e-06,
+      "loss": 0.0199,
+      "num_input_tokens_seen": 37070832,
+      "step": 903
+    },
+    {
+      "epoch": 16.440366972477065,
+      "grad_norm": 0.22360800206661224,
+      "learning_rate": 3.205389657580943e-06,
+      "loss": 0.021,
+      "num_input_tokens_seen": 37103136,
+      "step": 904
+    },
+    {
+      "epoch": 16.458715596330276,
+      "grad_norm": 0.22224462032318115,
+      "learning_rate": 3.1698561306951064e-06,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 37148584,
+      "step": 905
+    },
+    {
+      "epoch": 16.477064220183486,
+      "grad_norm": 0.26315537095069885,
+      "learning_rate": 3.1345073215151066e-06,
+      "loss": 0.021,
+      "num_input_tokens_seen": 37187096,
+      "step": 906
+    },
+    {
+      "epoch": 16.495412844036696,
+      "grad_norm": 0.26664742827415466,
+      "learning_rate": 3.0993435291480355e-06,
+      "loss": 0.0201,
+      "num_input_tokens_seen": 37225824,
+      "step": 907
+    },
+    {
+      "epoch": 16.513761467889907,
+      "grad_norm": 0.24455232918262482,
+      "learning_rate": 3.0643650511354484e-06,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 37262480,
+      "step": 908
+    },
+    {
+      "epoch": 16.53211009174312,
+      "grad_norm": 0.19751450419425964,
+      "learning_rate": 3.029572183450868e-06,
+      "loss": 0.0173,
+      "num_input_tokens_seen": 37303856,
+      "step": 909
+    },
+    {
+      "epoch": 16.55045871559633,
+      "grad_norm": 0.24461263418197632,
+      "learning_rate": 2.9949652204972254e-06,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 37342760,
+      "step": 910
+    },
+    {
+      "epoch": 16.568807339449542,
+      "grad_norm": 0.20236024260520935,
+      "learning_rate": 2.9605444551044166e-06,
+      "loss": 0.018,
+      "num_input_tokens_seen": 37388064,
+      "step": 911
+    },
+    {
+      "epoch": 16.587155963302752,
+      "grad_norm": 0.19530493021011353,
+      "learning_rate": 2.9263101785268254e-06,
+      "loss": 0.0153,
+      "num_input_tokens_seen": 37428768,
+      "step": 912
+    },
+    {
+      "epoch": 16.605504587155963,
+      "grad_norm": 0.21812154352664948,
+      "learning_rate": 2.89226268044083e-06,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 37474624,
+      "step": 913
+    },
+    {
+      "epoch": 16.623853211009173,
+      "grad_norm": 0.22331184148788452,
+      "learning_rate": 2.858402248942349e-06,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 37511824,
+      "step": 914
+    },
+    {
+      "epoch": 16.642201834862384,
+      "grad_norm": 0.2519419491291046,
+      "learning_rate": 2.8247291705444575e-06,
+      "loss": 0.0191,
+      "num_input_tokens_seen": 37550216,
+      "step": 915
+    },
+    {
+      "epoch": 16.660550458715598,
+      "grad_norm": 0.21349124610424042,
+      "learning_rate": 2.7912437301749026e-06,
+      "loss": 0.017,
+      "num_input_tokens_seen": 37600240,
+      "step": 916
+    },
+    {
+      "epoch": 16.678899082568808,
+      "grad_norm": 0.2626224756240845,
+      "learning_rate": 2.7579462111737065e-06,
+      "loss": 0.0231,
+      "num_input_tokens_seen": 37642608,
+      "step": 917
+    },
+    {
+      "epoch": 16.69724770642202,
+      "grad_norm": 0.28207895159721375,
+      "learning_rate": 2.7248368952908053e-06,
+      "loss": 0.0227,
+      "num_input_tokens_seen": 37684840,
+      "step": 918
+    },
+    {
+      "epoch": 16.71559633027523,
+      "grad_norm": 0.23850199580192566,
+      "learning_rate": 2.6919160626836166e-06,
+      "loss": 0.0167,
+      "num_input_tokens_seen": 37718968,
+      "step": 919
+    },
+    {
+      "epoch": 16.73394495412844,
+      "grad_norm": 0.26844337582588196,
+      "learning_rate": 2.659183991914696e-06,
+      "loss": 0.0307,
+      "num_input_tokens_seen": 37767744,
+      "step": 920
+    },
+    {
+      "epoch": 16.75229357798165,
+      "grad_norm": 0.23936662077903748,
+      "learning_rate": 2.6266409599493753e-06,
+      "loss": 0.0219,
+      "num_input_tokens_seen": 37815432,
+      "step": 921
+    },
+    {
+      "epoch": 16.770642201834864,
+      "grad_norm": 0.24459759891033173,
+      "learning_rate": 2.5942872421534146e-06,
+      "loss": 0.0175,
+      "num_input_tokens_seen": 37859656,
+      "step": 922
+    },
+    {
+      "epoch": 16.788990825688074,
+      "grad_norm": 0.2952236235141754,
+      "learning_rate": 2.5621231122906873e-06,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 37891816,
+      "step": 923
+    },
+    {
+      "epoch": 16.807339449541285,
+      "grad_norm": 0.2381288856267929,
+      "learning_rate": 2.5301488425208296e-06,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 37930184,
+      "step": 924
+    },
+    {
+      "epoch": 16.825688073394495,
+      "grad_norm": 0.28937825560569763,
+      "learning_rate": 2.4983647033969714e-06,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 37982736,
+      "step": 925
+    },
+    {
+      "epoch": 16.844036697247706,
+      "grad_norm": 0.21282249689102173,
+      "learning_rate": 2.4667709638634434e-06,
+      "loss": 0.0177,
+      "num_input_tokens_seen": 38022992,
+      "step": 926
+    },
+    {
+      "epoch": 16.862385321100916,
+      "grad_norm": 0.24419566988945007,
+      "learning_rate": 2.43536789125349e-06,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 38070776,
+      "step": 927
+    },
+    {
+      "epoch": 16.88073394495413,
+      "grad_norm": 0.2718816101551056,
+      "learning_rate": 2.4041557512869878e-06,
+      "loss": 0.0266,
+      "num_input_tokens_seen": 38122544,
+      "step": 928
+    },
+    {
+      "epoch": 16.89908256880734,
+      "grad_norm": 0.2344757467508316,
+      "learning_rate": 2.3731348080682504e-06,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 38160680,
+      "step": 929
+    },
+    {
+      "epoch": 16.91743119266055,
+      "grad_norm": 0.2165100872516632,
+      "learning_rate": 2.3423053240837515e-06,
+      "loss": 0.0176,
+      "num_input_tokens_seen": 38200720,
+      "step": 930
+    },
+    {
+      "epoch": 16.93577981651376,
+      "grad_norm": 0.24561312794685364,
+      "learning_rate": 2.3116675601999094e-06,
+      "loss": 0.0196,
+      "num_input_tokens_seen": 38240296,
+      "step": 931
+    },
+    {
+      "epoch": 16.954128440366972,
+      "grad_norm": 0.2633242607116699,
+      "learning_rate": 2.281221775660894e-06,
+      "loss": 0.0275,
+      "num_input_tokens_seen": 38299672,
+      "step": 932
+    },
+    {
+      "epoch": 16.972477064220183,
+      "grad_norm": 0.21266628801822662,
+      "learning_rate": 2.2509682280864224e-06,
+      "loss": 0.02,
+      "num_input_tokens_seen": 38350408,
+      "step": 933
+    },
+    {
+      "epoch": 16.990825688073393,
+      "grad_norm": 0.22149045765399933,
+      "learning_rate": 2.2209071734695746e-06,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 38386312,
+      "step": 934
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 0.3346998989582062,
+      "learning_rate": 2.1910388661746493e-06,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 38404496,
+      "step": 935
+    },
+    {
+      "epoch": 17.01834862385321,
+      "grad_norm": 0.24249973893165588,
+      "learning_rate": 2.1613635589349756e-06,
+      "loss": 0.0216,
+      "num_input_tokens_seen": 38440632,
+      "step": 936
+    },
+    {
+      "epoch": 17.03669724770642,
+      "grad_norm": 0.2142462283372879,
+      "learning_rate": 2.131881502850824e-06,
+      "loss": 0.0149,
+      "num_input_tokens_seen": 38480008,
+      "step": 937
+    },
+    {
+      "epoch": 17.05504587155963,
+      "grad_norm": 0.22429676353931427,
+      "learning_rate": 2.1025929473872274e-06,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 38523112,
+      "step": 938
+    },
+    {
+      "epoch": 17.073394495412845,
+      "grad_norm": 0.22950181365013123,
+      "learning_rate": 2.073498140371899e-06,
+      "loss": 0.025,
+      "num_input_tokens_seen": 38564512,
+      "step": 939
+    },
+    {
+      "epoch": 17.091743119266056,
+      "grad_norm": 0.33720073103904724,
+      "learning_rate": 2.044597327993153e-06,
+      "loss": 0.0269,
+      "num_input_tokens_seen": 38595896,
+      "step": 940
+    },
+    {
+      "epoch": 17.110091743119266,
+      "grad_norm": 0.22238488495349884,
+      "learning_rate": 2.0158907547977722e-06,
+      "loss": 0.0183,
+      "num_input_tokens_seen": 38628184,
+      "step": 941
+    },
+    {
+      "epoch": 17.128440366972477,
+      "grad_norm": 0.29164692759513855,
+      "learning_rate": 1.9873786636889906e-06,
+      "loss": 0.0283,
+      "num_input_tokens_seen": 38667296,
+      "step": 942
+    },
+    {
+      "epoch": 17.146788990825687,
+      "grad_norm": 0.1883544772863388,
+      "learning_rate": 1.9590612959244055e-06,
+      "loss": 0.0135,
+      "num_input_tokens_seen": 38707024,
+      "step": 943
+    },
+    {
+      "epoch": 17.165137614678898,
+      "grad_norm": 0.25679290294647217,
+      "learning_rate": 1.9309388911139426e-06,
+      "loss": 0.0242,
+      "num_input_tokens_seen": 38748056,
+      "step": 944
+    },
+    {
+      "epoch": 17.18348623853211,
+      "grad_norm": 0.20496375858783722,
+      "learning_rate": 1.9030116872178316e-06,
+      "loss": 0.0166,
+      "num_input_tokens_seen": 38786744,
+      "step": 945
+    },
+    {
+      "epoch": 17.201834862385322,
+      "grad_norm": 0.23621124029159546,
+      "learning_rate": 1.8752799205445982e-06,
+      "loss": 0.0219,
+      "num_input_tokens_seen": 38840712,
+      "step": 946
+    },
+    {
+      "epoch": 17.220183486238533,
+      "grad_norm": 0.259958416223526,
+      "learning_rate": 1.8477438257490543e-06,
+      "loss": 0.0189,
+      "num_input_tokens_seen": 38871016,
+      "step": 947
+    },
+    {
+      "epoch": 17.238532110091743,
+      "grad_norm": 0.19008736312389374,
+      "learning_rate": 1.8204036358303173e-06,
+      "loss": 0.0141,
+      "num_input_tokens_seen": 38919976,
+      "step": 948
+    },
+    {
+      "epoch": 17.256880733944953,
+      "grad_norm": 0.15731441974639893,
+      "learning_rate": 1.793259582129836e-06,
+      "loss": 0.0145,
+      "num_input_tokens_seen": 38973000,
+      "step": 949
+    },
+    {
+      "epoch": 17.275229357798164,
+      "grad_norm": 0.20690017938613892,
+      "learning_rate": 1.7663118943294366e-06,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 39027080,
+      "step": 950
+    },
+    {
+      "epoch": 17.293577981651374,
+      "grad_norm": 0.27831098437309265,
+      "learning_rate": 1.7395608004493886e-06,
+      "loss": 0.0214,
+      "num_input_tokens_seen": 39055792,
+      "step": 951
+    },
+    {
+      "epoch": 17.31192660550459,
+      "grad_norm": 0.23228834569454193,
+      "learning_rate": 1.713006526846439e-06,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 39108640,
+      "step": 952
+    },
+    {
+      "epoch": 17.3302752293578,
+      "grad_norm": 0.2643783688545227,
+      "learning_rate": 1.686649298211951e-06,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 39152080,
+      "step": 953
+    },
+    {
+      "epoch": 17.34862385321101,
+      "grad_norm": 0.18884579837322235,
+      "learning_rate": 1.6604893375699594e-06,
+      "loss": 0.0148,
+      "num_input_tokens_seen": 39199992,
+      "step": 954
+    },
+    {
+      "epoch": 17.36697247706422,
+      "grad_norm": 0.2350739687681198,
+      "learning_rate": 1.6345268662752906e-06,
+      "loss": 0.019,
+      "num_input_tokens_seen": 39239008,
+      "step": 955
+    },
+    {
+      "epoch": 17.38532110091743,
+      "grad_norm": 0.20768049359321594,
+      "learning_rate": 1.6087621040117157e-06,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 39274760,
+      "step": 956
+    },
+    {
+      "epoch": 17.40366972477064,
+      "grad_norm": 0.23328879475593567,
+      "learning_rate": 1.5831952687900608e-06,
+      "loss": 0.0191,
+      "num_input_tokens_seen": 39306592,
+      "step": 957
+    },
+    {
+      "epoch": 17.422018348623855,
+      "grad_norm": 0.2595842778682709,
+      "learning_rate": 1.5578265769463806e-06,
+      "loss": 0.0192,
+      "num_input_tokens_seen": 39343144,
+      "step": 958
+    },
+    {
+      "epoch": 17.440366972477065,
+      "grad_norm": 0.2666177451610565,
+      "learning_rate": 1.5326562431401153e-06,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 39381656,
+      "step": 959
+    },
+    {
+      "epoch": 17.458715596330276,
+      "grad_norm": 0.2561241686344147,
+      "learning_rate": 1.5076844803522922e-06,
+      "loss": 0.021,
+      "num_input_tokens_seen": 39421896,
+      "step": 960
+    },
+    {
+      "epoch": 17.477064220183486,
+      "grad_norm": 0.24035242199897766,
+      "learning_rate": 1.482911499883699e-06,
+      "loss": 0.0214,
+      "num_input_tokens_seen": 39459352,
+      "step": 961
+    },
+    {
+      "epoch": 17.495412844036696,
+      "grad_norm": 0.2297283411026001,
+      "learning_rate": 1.4583375113531195e-06,
+      "loss": 0.0146,
+      "num_input_tokens_seen": 39492800,
+      "step": 962
+    },
+    {
+      "epoch": 17.513761467889907,
+      "grad_norm": 0.19542022049427032,
+      "learning_rate": 1.4339627226955392e-06,
+      "loss": 0.0165,
+      "num_input_tokens_seen": 39537000,
+      "step": 963
+    },
+    {
+      "epoch": 17.53211009174312,
+      "grad_norm": 0.2531075179576874,
+      "learning_rate": 1.4097873401604122e-06,
+      "loss": 0.0232,
+      "num_input_tokens_seen": 39576288,
+      "step": 964
+    },
+    {
+      "epoch": 17.55045871559633,
+      "grad_norm": 0.20797313749790192,
+      "learning_rate": 1.3858115683098832e-06,
+      "loss": 0.0142,
+      "num_input_tokens_seen": 39612936,
+      "step": 965
+    },
+    {
+      "epoch": 17.568807339449542,
+      "grad_norm": 0.23901662230491638,
+      "learning_rate": 1.362035610017079e-06,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 39651632,
+      "step": 966
+    },
+    {
+      "epoch": 17.587155963302752,
+      "grad_norm": 0.23252394795417786,
+      "learning_rate": 1.3384596664643922e-06,
+      "loss": 0.0201,
+      "num_input_tokens_seen": 39713824,
+      "step": 967
+    },
+    {
+      "epoch": 17.605504587155963,
+      "grad_norm": 0.21391186118125916,
+      "learning_rate": 1.31508393714177e-06,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 39753616,
+      "step": 968
+    },
+    {
+      "epoch": 17.623853211009173,
+      "grad_norm": 0.2858409583568573,
+      "learning_rate": 1.291908619845017e-06,
+      "loss": 0.0284,
+      "num_input_tokens_seen": 39789208,
+      "step": 969
+    },
+    {
+      "epoch": 17.642201834862384,
+      "grad_norm": 0.23830647766590118,
+      "learning_rate": 1.2689339106741527e-06,
+      "loss": 0.0219,
+      "num_input_tokens_seen": 39823072,
+      "step": 970
+    },
+    {
+      "epoch": 17.660550458715598,
+      "grad_norm": 0.2422226369380951,
+      "learning_rate": 1.2461600040317212e-06,
+      "loss": 0.0221,
+      "num_input_tokens_seen": 39853600,
+      "step": 971
+    },
+    {
+      "epoch": 17.678899082568808,
+      "grad_norm": 0.22980886697769165,
+      "learning_rate": 1.2235870926211619e-06,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 39890080,
+      "step": 972
+    },
+    {
+      "epoch": 17.69724770642202,
+      "grad_norm": 0.2705008089542389,
+      "learning_rate": 1.2012153674451715e-06,
+      "loss": 0.0258,
+      "num_input_tokens_seen": 39929208,
+      "step": 973
+    },
+    {
+      "epoch": 17.71559633027523,
+      "grad_norm": 0.25133201479911804,
+      "learning_rate": 1.1790450178040951e-06,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 39969344,
+      "step": 974
+    },
+    {
+      "epoch": 17.73394495412844,
+      "grad_norm": 0.2319001406431198,
+      "learning_rate": 1.1570762312943295e-06,
+      "loss": 0.0188,
+      "num_input_tokens_seen": 40012816,
+      "step": 975
+    },
+    {
+      "epoch": 17.75229357798165,
+      "grad_norm": 0.27038779854774475,
+      "learning_rate": 1.1353091938067023e-06,
+      "loss": 0.0215,
+      "num_input_tokens_seen": 40051096,
+      "step": 976
+    },
+    {
+      "epoch": 17.770642201834864,
+      "grad_norm": 0.24062982201576233,
+      "learning_rate": 1.1137440895249457e-06,
+      "loss": 0.0173,
+      "num_input_tokens_seen": 40090488,
+      "step": 977
+    },
+    {
+      "epoch": 17.788990825688074,
+      "grad_norm": 0.202128067612648,
+      "learning_rate": 1.0923811009241142e-06,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 40145176,
+      "step": 978
+    },
+    {
+      "epoch": 17.807339449541285,
+      "grad_norm": 0.2674156129360199,
+      "learning_rate": 1.071220408769033e-06,
+      "loss": 0.0246,
+      "num_input_tokens_seen": 40189416,
+      "step": 979
+    },
+    {
+      "epoch": 17.825688073394495,
+      "grad_norm": 0.1906251758337021,
+      "learning_rate": 1.0502621921127776e-06,
+      "loss": 0.014,
+      "num_input_tokens_seen": 40239272,
+      "step": 980
+    },
+    {
+      "epoch": 17.844036697247706,
+      "grad_norm": 0.23186717927455902,
+      "learning_rate": 1.0295066282951738e-06,
+      "loss": 0.0239,
+      "num_input_tokens_seen": 40288104,
+      "step": 981
+    },
+    {
+      "epoch": 17.862385321100916,
+      "grad_norm": 0.2101387232542038,
+      "learning_rate": 1.0089538929412724e-06,
+      "loss": 0.0168,
+      "num_input_tokens_seen": 40322904,
+      "step": 982
+    },
+    {
+      "epoch": 17.88073394495413,
+      "grad_norm": 0.23706017434597015,
+      "learning_rate": 9.886041599598606e-07,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 40372536,
+      "step": 983
+    },
+    {
+      "epoch": 17.89908256880734,
+      "grad_norm": 0.21513932943344116,
+      "learning_rate": 9.684576015420278e-07,
+      "loss": 0.0188,
+      "num_input_tokens_seen": 40427408,
+      "step": 984
+    },
+    {
+      "epoch": 17.91743119266055,
+      "grad_norm": 0.21823860704898834,
+      "learning_rate": 9.485143881596714e-07,
+      "loss": 0.0238,
+      "num_input_tokens_seen": 40477568,
+      "step": 985
+    },
+    {
+      "epoch": 17.93577981651376,
+      "grad_norm": 0.24832789599895477,
+      "learning_rate": 9.287746885640603e-07,
+      "loss": 0.0215,
+      "num_input_tokens_seen": 40519224,
+      "step": 986
+    },
+    {
+      "epoch": 17.954128440366972,
+      "grad_norm": 0.2347378134727478,
+      "learning_rate": 9.092386697844263e-07,
+      "loss": 0.0188,
+      "num_input_tokens_seen": 40560640,
+      "step": 987
+    },
+    {
+      "epoch": 17.972477064220183,
+      "grad_norm": 0.26193633675575256,
+      "learning_rate": 8.899064971265276e-07,
+      "loss": 0.029,
+      "num_input_tokens_seen": 40609544,
+      "step": 988
+    },
+    {
+      "epoch": 17.990825688073393,
+      "grad_norm": 0.21652978658676147,
+      "learning_rate": 8.707783341712656e-07,
+      "loss": 0.0173,
+      "num_input_tokens_seen": 40646216,
+      "step": 989
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 0.3287041187286377,
+      "learning_rate": 8.51854342773295e-07,
+      "loss": 0.0252,
+      "num_input_tokens_seen": 40663584,
+      "step": 990
+    },
+    {
+      "epoch": 18.01834862385321,
+      "grad_norm": 0.20799626410007477,
+      "learning_rate": 8.331346830596498e-07,
+      "loss": 0.0136,
+      "num_input_tokens_seen": 40711208,
+      "step": 991
+    },
+    {
+      "epoch": 18.03669724770642,
+      "grad_norm": 0.21691110730171204,
+      "learning_rate": 8.146195134284052e-07,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 40752144,
+      "step": 992
+    },
+    {
+      "epoch": 18.05504587155963,
+      "grad_norm": 0.23775102198123932,
+      "learning_rate": 7.963089905473092e-07,
+      "loss": 0.0204,
+      "num_input_tokens_seen": 40798560,
+      "step": 993
+    },
+    {
+      "epoch": 18.073394495412845,
+      "grad_norm": 0.23806151747703552,
+      "learning_rate": 7.782032693524816e-07,
+      "loss": 0.0196,
+      "num_input_tokens_seen": 40829872,
+      "step": 994
+    },
+    {
+      "epoch": 18.091743119266056,
+      "grad_norm": 0.2555314004421234,
+      "learning_rate": 7.603025030471e-07,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 40866712,
+      "step": 995
+    },
+    {
+      "epoch": 18.110091743119266,
+      "grad_norm": 0.20765069127082825,
+      "learning_rate": 7.426068431000882e-07,
+      "loss": 0.0152,
+      "num_input_tokens_seen": 40917200,
+      "step": 996
+    },
+    {
+      "epoch": 18.128440366972477,
+      "grad_norm": 0.2072562575340271,
+      "learning_rate": 7.251164392448496e-07,
+      "loss": 0.0176,
+      "num_input_tokens_seen": 40961592,
+      "step": 997
+    },
+    {
+      "epoch": 18.146788990825687,
+      "grad_norm": 0.24391891062259674,
+      "learning_rate": 7.078314394779961e-07,
+      "loss": 0.0215,
+      "num_input_tokens_seen": 40996304,
+      "step": 998
+    },
+    {
+      "epoch": 18.165137614678898,
+      "grad_norm": 0.21016541123390198,
+      "learning_rate": 6.907519900580861e-07,
+      "loss": 0.0196,
+      "num_input_tokens_seen": 41054400,
+      "step": 999
+    },
+    {
+      "epoch": 18.18348623853211,
+      "grad_norm": 0.2642466723918915,
+      "learning_rate": 6.738782355044049e-07,
+      "loss": 0.0189,
+      "num_input_tokens_seen": 41106064,
+      "step": 1000
+    },
+    {
+      "epoch": 18.201834862385322,
+      "grad_norm": 0.24111659824848175,
+      "learning_rate": 6.572103185957301e-07,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 41148920,
+      "step": 1001
+    },
+    {
+      "epoch": 18.220183486238533,
+      "grad_norm": 0.2610417604446411,
+      "learning_rate": 6.407483803691216e-07,
+      "loss": 0.022,
+      "num_input_tokens_seen": 41176536,
+      "step": 1002
+    },
+    {
+      "epoch": 18.238532110091743,
+      "grad_norm": 0.2215040922164917,
+      "learning_rate": 6.244925601187363e-07,
+      "loss": 0.0183,
+      "num_input_tokens_seen": 41223776,
+      "step": 1003
+    },
+    {
+      "epoch": 18.256880733944953,
+      "grad_norm": 0.19494834542274475,
+      "learning_rate": 6.084429953946452e-07,
+      "loss": 0.0182,
+      "num_input_tokens_seen": 41274976,
+      "step": 1004
+    },
+    {
+      "epoch": 18.275229357798164,
+      "grad_norm": 0.314108669757843,
+      "learning_rate": 5.925998220016659e-07,
+      "loss": 0.0273,
+      "num_input_tokens_seen": 41300896,
+      "step": 1005
+    },
+    {
+      "epoch": 18.293577981651374,
+      "grad_norm": 0.22374023497104645,
+      "learning_rate": 5.769631739982267e-07,
+      "loss": 0.0153,
+      "num_input_tokens_seen": 41338840,
+      "step": 1006
+    },
+    {
+      "epoch": 18.31192660550459,
+      "grad_norm": 0.21842226386070251,
+      "learning_rate": 5.615331836952121e-07,
+      "loss": 0.0172,
+      "num_input_tokens_seen": 41374664,
+      "step": 1007
+    },
+    {
+      "epoch": 18.3302752293578,
+      "grad_norm": 0.17510443925857544,
+      "learning_rate": 5.463099816548579e-07,
+      "loss": 0.0141,
+      "num_input_tokens_seen": 41413920,
+      "step": 1008
+    },
+    {
+      "epoch": 18.34862385321101,
+      "grad_norm": 0.3019046485424042,
+      "learning_rate": 5.312936966896498e-07,
+      "loss": 0.0343,
+      "num_input_tokens_seen": 41461272,
+      "step": 1009
+    },
+    {
+      "epoch": 18.36697247706422,
+      "grad_norm": 0.24767573177814484,
+      "learning_rate": 5.164844558612131e-07,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 41501328,
+      "step": 1010
+    },
+    {
+      "epoch": 18.38532110091743,
+      "grad_norm": 0.2388239949941635,
+      "learning_rate": 5.018823844792603e-07,
+      "loss": 0.0208,
+      "num_input_tokens_seen": 41534096,
+      "step": 1011
+    },
+    {
+      "epoch": 18.40366972477064,
+      "grad_norm": 0.23209671676158905,
+      "learning_rate": 4.874876061005173e-07,
+      "loss": 0.0189,
+      "num_input_tokens_seen": 41576592,
+      "step": 1012
+    },
+    {
+      "epoch": 18.422018348623855,
+      "grad_norm": 0.1739690899848938,
+      "learning_rate": 4.7330024252768555e-07,
+      "loss": 0.0128,
+      "num_input_tokens_seen": 41624000,
+      "step": 1013
+    },
+    {
+      "epoch": 18.440366972477065,
+      "grad_norm": 0.22390468418598175,
+      "learning_rate": 4.5932041380840065e-07,
+      "loss": 0.0231,
+      "num_input_tokens_seen": 41656024,
+      "step": 1014
+    },
+    {
+      "epoch": 18.458715596330276,
+      "grad_norm": 0.2131803333759308,
+      "learning_rate": 4.4554823823423354e-07,
+      "loss": 0.0154,
+      "num_input_tokens_seen": 41697984,
+      "step": 1015
+    },
+    {
+      "epoch": 18.477064220183486,
+      "grad_norm": 0.22676204144954681,
+      "learning_rate": 4.319838323396691e-07,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 41740560,
+      "step": 1016
+    },
+    {
+      "epoch": 18.495412844036696,
+      "grad_norm": 0.1953912377357483,
+      "learning_rate": 4.1862731090113736e-07,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 41783512,
+      "step": 1017
+    },
+    {
+      "epoch": 18.513761467889907,
+      "grad_norm": 0.21998439729213715,
+      "learning_rate": 4.05478786936031e-07,
+      "loss": 0.018,
+      "num_input_tokens_seen": 41822144,
+      "step": 1018
+    },
+    {
+      "epoch": 18.53211009174312,
+      "grad_norm": 0.2555588483810425,
+      "learning_rate": 3.9253837170175887e-07,
+      "loss": 0.0157,
+      "num_input_tokens_seen": 41860344,
+      "step": 1019
+    },
+    {
+      "epoch": 18.55045871559633,
+      "grad_norm": 0.2640567719936371,
+      "learning_rate": 3.7980617469479953e-07,
+      "loss": 0.0278,
+      "num_input_tokens_seen": 41915536,
+      "step": 1020
+    },
+    {
+      "epoch": 18.568807339449542,
+      "grad_norm": 0.21900391578674316,
+      "learning_rate": 3.672823036497658e-07,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 41961056,
+      "step": 1021
+    },
+    {
+      "epoch": 18.587155963302752,
+      "grad_norm": 0.23652507364749908,
+      "learning_rate": 3.5496686453850846e-07,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 41998240,
+      "step": 1022
+    },
+    {
+      "epoch": 18.605504587155963,
+      "grad_norm": 0.23997560143470764,
+      "learning_rate": 3.428599615692141e-07,
+      "loss": 0.0176,
+      "num_input_tokens_seen": 42032136,
+      "step": 1023
+    },
+    {
+      "epoch": 18.623853211009173,
+      "grad_norm": 0.22659312188625336,
+      "learning_rate": 3.309616971855195e-07,
+      "loss": 0.0185,
+      "num_input_tokens_seen": 42079960,
+      "step": 1024
+    },
+    {
+      "epoch": 18.642201834862384,
+      "grad_norm": 0.1882646530866623,
+      "learning_rate": 3.192721720656489e-07,
+      "loss": 0.0148,
+      "num_input_tokens_seen": 42120432,
+      "step": 1025
+    },
+    {
+      "epoch": 18.660550458715598,
+      "grad_norm": 0.20917965471744537,
+      "learning_rate": 3.077914851215585e-07,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 42171464,
+      "step": 1026
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1080,
+  "num_input_tokens_seen": 42171464,
+  "num_train_epochs": 20,
+  "save_steps": 54,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.5576552468869284e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}